Back to the table

79 Beginner Cx Climax Plot & Beats

Climax

The final confrontation that answers the story's question.

What it is

The climax is the scene where the question your novel has been asking finally gets answered. Every novel asks one. Somewhere in the first fifty pages the reader starts to wonder about one specific thing: will this woman get her daughter back, will this detective name the killer, will these two end up together, will this city fall. At the climax your protagonist (your main character) meets the main opposition head-on, and the answer is fixed for good. Nothing after it can reopen the question.

Each part of that sentence is doing a job, so take them in turn. The main opposition is more than just another obstacle. A book may throw thirty obstacles at its protagonist, but one thing has been blocking her all along, and that thing might be a person, an institution, a system, a stretch of wild country or her own firm belief. At the climax she faces it directly instead of working around it. Antagonist covers what makes the opposition able to show up at all. Then the answer. The climax is where the question gets its yes, its no, or a version of either that comes with a price, and that isn't always where the loudest thing in the book happens. The question itself belongs to Dramatic Question, and this page picks up one step later, so if you can't state yours in one sentence, read that page first. And the climax is action. Whatever the protagonist has learned across the book shows up here because she does something with it. A character who explains her change in dialogue hasn't had a climax, just a conversation about one.

If you take one thing from this page, take this. The protagonist wins the climax, and she wins it with something the story has paid for. Everything else here fills that sentence out, so look at its two halves.

She wins it. That doesn't mean she gets what she wanted, because a climax can be lost, and losing gets its own treatment below. It means the deciding act is hers. She's the subject of the verb that settles the question. If the mentor rides in with the cavalry, or the antagonist has a change of heart, or a storm sinks the ship, then the three hundred pages the reader spent caring about this person have quietly been taken back. Agency counts for more here than anywhere else in the novel, because this is where the reader is openly watching to see whether it was all going somewhere. Agency covers her power to act across the whole book, so go there if the problem turns out to run through the whole draft.

With something the story paid for. She has to have bought the means earlier, at a price, in a scene the reader remembers: a skill we saw her practicing, a fact we saw her get wrong first, a relationship we saw her repair, a weakness we saw her refuse to look at. Readers call that "earned", and its absence is what they mean by "contrived", "convenient" or "out of nowhere". None of this means the reader has to guess it. The best version is one the reader didn't see coming, and the moment it happens they recognize every piece of it. William Archer, whose 1912 manual on playwriting is one of this page's examples, had a phrase for the trick behind that effect: "foreshadowing without forestalling."

How "paid for" works, and the many ways an ending can cheat, has a whole neighboring element to itself, including the oldest cheat of all, a god lowered onto the stage by a machine. The point from it that matters most here is that luck is rationed by where it falls in the book, and the climax is where the ration runs out. Cause and Effect owns that ground, deus ex machina included, along with the question of how much coincidence a story can take, and where. Read it before you try to work out what's wrong with your own third act.

The theater has a second way of describing the climax, and some writers find it more useful than any definition based on structure. An audience watching a play starts to expect things, and they expect more than a good play. They expect particular scenes. Once the son learns that the man he's been asked to rescue is his father, the audience knows there'll be a scene between those two men, and they'd feel cheated by a play that skipped it. French critics in the 1800s had a name for that kind of scene, the scène à faire, the scene that must be done, and Archer put it into English as the obligatory scene. Writers have used the term ever since. What it tells you is that your climax isn't free. You chose it in your first act, at least in outline, through the promises the book made. You can surprise the reader with how the scene goes, but you can't swap in a different scene and expect them to be happy. So a climax can be well made and still feel wrong, and the writing isn't the problem. The scene is answering a question nobody asked.

Of all the faults in third acts, that one gets missed most often, so here's how it happens. A writer sets up a book about a woman trying to get her brother out of prison. Forty thousand words in, the prison plot has gotten hard to write and the romance subplot has gotten fun, so the writer follows the fun. The climax, when it comes, is a superb scene in which she chooses the wrong man and knows it. Readers finish the book with the odd feeling of having been handed something they didn't order. The brother is still in prison, and not because the ending was meant to be bleak. The book just forgot he was the reason anyone started reading.

In commercial fiction this idea belongs to genre, and Shawn Coyne pushed it furthest in The Story Grid (Black Irish Entertainment, 2015). He treats each genre as owing its readers a set of conventions and obligatory scenes. In a murder mystery, those are the finding of the body, the face-off with the killer and the delivery of a verdict. Story Grid covers the method itself, and Reader Promise covers the wider bargain your cover, blurb and category make for you. Both come back to the same idea. The ending was promised early, and the climax is where the reader collects.

Now, if your ending feels flat even though a lot happens in it, this next part is probably the reason. Most novels run two lines at once. The outer line is the prison, the killer, the siege or the deal. The inner line is the character changing, or failing to. At the climax both lines come due. The outer climax is the confrontation, and the inner climax is the choice she makes inside it. Character Arc owns the inner line across the book, and Want vs Need owns the gap between what she's chasing and what would actually help her.

When the two happen in the same moment, the ending lands hard, because one action does two jobs. She does the thing that settles the plot, and only the person she's become could do it. Most craft books mean exactly that when they say a climax should "prove" the arc, and Jill Chamberlain's The Nutshell Technique (University of Texas Press, 2016) builds a whole method around the join. She calls it the Climactic Choice and treats it as the decision that has to carry the protagonist from the flaw set up at the start to the strength shown at the end.

When they're separate scenes, the order matters, because it tells the reader what kind of book this was. Inner first, then outer, reads as a story about a person. She decides who she is, then goes and does the hard thing, and the doing shows us what she already found rather than finding something new. Outer first, then inner, reads as a story about what things cost. The fight happens, she survives it or she doesn't, and afterward, in a much quieter scene, she makes the choice that shows what it cost her. Both orders work. What doesn't work is leaving one of them out, or letting a long stretch of aftermath sit between them, because once the outer question is settled the reader's attention shuts down, and anything in that gap feels like drag.

Rocky (1976) puts them in the same moment. Rocky's stated goal isn't to win but to go the distance, so the final bell settles the outer question by losing it and the inner one by winning it, in the same second. Martha Wells's Murderbot novella All Systems Red, which comes up again below, splits them and puts the inner one last. The plot is resolved and the survivors are safe, and then, pages later, the protagonist makes a choice about what it wants to be.

So when an ending feels flat, ask a short question. Which of the two do you have? Most flat endings have a solid outer climax and no inner one, which is why they feel busy but don't move anyone. Fewer have an inner climax with no outer one, and readers say those books trail off, or that nothing much happens.

You'll often hear that the climax is the peak of tension, and that's true. The same sentence is also the most misleading one in all the advice on structure, because writers read "peak" as "loudest". Tension is the reader wanting to know and being made to wait. It depends on how much is at risk and how unsure the outcome is, and how much is happening doesn't come into it. A car chase has no tension if the reader is sure the hero will live, while a woman deciding whether to sign a piece of paper has huge tension if the reader knows what signing will cost her and really doesn't know which way she'll go. The peak, then, is the point where the most is at stake and the reader is least sure how it'll go. The peak is about doubt, and noise has nothing to do with it.

Three practical things follow from that. A quiet climax can be the strongest kind, because it strips away the noise, the crowd and the moving parts and leaves the reader alone with a decision, and when a book is really about what goes on inside a person, the quiet climax is more than a matter of style, because it's the only version that answers the right question. There's a full example below. Next, the outcome has to still be in doubt when you get there. A lot of climaxes fall flat because the answer became obvious sixty pages earlier and the writer kept raising the pressure anyway, and pressure without doubt just makes noise. See Rising Action for what real escalation is made of: fewer and fewer moves left open to the character, rather than a volume knob turned up. And the protagonist has to be able to lose. If the reader can't picture a believable version of this scene where she loses, the scene has no tension, whatever is blowing up in it. Stakes covers how to make the losing version believable.

Where does it sit? In a typical commercial novel the climax comes late, after the third-act low point and before the wind-down, and everything around it has its own page on this table: All Is Lost for the collapse just before it, Midpoint for the hinge in the middle, Resolution for what comes after, and Three-Act Structure for the frame they all sit in. The climax isn't the end of the book, but it isn't far from it. K.M. Weiland's free breakdown of story structure has the climax starting around the 88 percent mark, with the single climactic moment as the next-to-last scene. Treat any percentage as a picture of what commercial fiction tends to do rather than a rule, but the shape behind it holds, because you want enough book left for the reader to land and not so much that they get bored. If you have eighty pages after your climax, you have two books. If you have none, you've stopped instead of ending. Resolution explains what those last pages are for, the aftermath, the settling, the new normal. This page hands all of that over and stops at the moment the question is answered.

The word itself hasn't always meant this. "Climax" came into talk about structure through Gustav Freytag's Die Technik des Dramas (1863), which Elias J. MacEwan translated into English in 1895, and for Freytag the climax is the middle of the play, the point where the protagonist seems most successful, after which the action falls away. Freytag's high point is much closer to what this table calls the Midpoint than to what most writers mean by the climax now, and Freytag's Pyramid covers his model on its own terms. People often cite Aristotle here too, and he has no word for the modern climax at all. The Poetics names the change of fortune, the reversal (peripeteia), the recognition (anagnorisis), and the point where the complication ends and the unraveling begins. It also insists that reversal and recognition have to grow out of the plot itself, as a necessary or likely result of what came before, instead of being forced in from outside. That rule is the ancestor of everything on this page, though the diagram with the peak in the middle isn't his.

One more distinction before the kinds of ending. A climactic scene and a climactic sequence are different animals. A single scene works when the question can be settled in one room, with a confession, a vote, a refusal or a signature. A sequence is several scenes in a row, sometimes across chapters, with no break between them, and you need one when the settling happens in stages: get in, get past the guard, find the room, face the person, deal with what follows. The difference matters for two reasons. A sequence needs one clear moment inside it where the question flips, with everything else in the sequence as lead-up and fallout, because without that moment readers can't tell they've reached the climax at all. And a sequence is where false climaxes breed, because each stage looks like a peak.

Most writing about the climax quietly assumes the hero wins, and that cuts writers off from whole kinds of ending. A climax can be lost and still satisfy completely, because the reader doesn't need the protagonist to win. They need the question answered, and they need her to be the one who answered it. Take "Does she get her daughter back?" A fully paid-for no is a satisfying climax. The question left unanswered isn't, and "she gets her daughter back because the antagonist relents" is worse than either. A lost climax can take three shapes, and each does something different.

The first is tragedy. The protagonist acts, the act is fully hers, and it destroys her. The satisfaction comes from the sense that it had to happen. Given who she is and what she's decided, this ending was always waiting, and the reader can see the whole chain at once. Notice what that asks of the climax. A tragic climax isn't a passive one, and in a good tragedy the protagonist is more active than in most comedies, because the disaster has to be something she does. Character Arc covers the downward arc in detail.

The second is the loss that is a win by a different measure. The outer question gets a no and the inner one gets a yes. She fails to save the company but saves her marriage, or loses the fight but stops being afraid, or doesn't get the promotion and stops wanting it. Commercial fiction uses this kind of lost climax more than any other, and the risk is that the outer loss gets treated as a footnote. If the reader spent the book wanting the company saved, losing it has to hurt on the page, or the ending feels like a trade.

The third is the want granted, and wrong. This one's the hardest, and often the best. The protagonist gets the very thing she's been chasing for four hundred pages, and getting it shows that it was the wrong thing all along. Here Want vs Need is at its most brutal, and the climax is how it works, because the moment she gets it is the moment it's exposed. The hard part is that you have to set up the want as something truly worth having, not an obvious mistake, or the reader gets there two hundred pages ahead of you. There's a well-known example of this below, and a shorter one here, though its last scene is the point of the film, so look away if you haven't seen it. Damien Chazelle's Whiplash (2014) ends with the drummer giving the performance he's wrecked his life to be able to give. He gets the approval he's been chasing, and the film won't tell you whether that's a triumph or a ruin.

One warning covers all three. A losing climax isn't a license to be bleak, and readers don't reward bleakness for its own sake. What they reward is being told the truth about the story they were given. If a book's premise, cover and first act promise a rescue and it delivers a tragedy instead, it has broken a promise, however good the tragedy is. Whether that promise holds is a Reader Promise question rather than a question of quality, and it's decided long before you write the climax.

Finally, four neighboring elements often get mixed up with this one. The Midpoint is the hinge in the middle, where the protagonist's understanding or position flips and the second half of the book becomes a different job from the first. In Freytag's terms it's the high point, but most working novelists don't mean the climax when they say that, and treating it as one gives you a book that peaks halfway through, at fifty percent. All Is Lost is the collapse before the climax, the opposite kind of scene, showing the protagonist at her worst, stripped of the means she was relying on. Martha Alderson insists on this split in The Plot Whisperer, where she calls the earlier scene the crisis and saves the word climax for the point where the protagonist acts from her strongest position, and she's right that mixing them up is what makes a third act read as two endings in a row. The Resolution is the aftermath, the new normal, the costs counted, the loose threads tied, the last image. Aftermath is a separate job with its own risks, and it has its own page. And a plot twist is a reveal that changes what the reader thought they knew. It can sit inside a climax, and in a mystery it usually does, but a twist is information and a climax is action. When a book's ending is only a reveal, it tends to get the review that says the ending was clever but didn't move anyone.

Why it matters

Readers use the ending to decide what the whole book was. Nobody remembers a novel evenly. They remember the opening, one or two scenes in the middle, and the ending, and the ending gets the final word on the rest. A weak climax lets them down on the last page and then reaches back and changes how they rate everything you did before it. So when a review says a book started well and then fell apart, nothing usually fell apart except the last forty pages.

If you publish on a schedule, the ending decides whether a reader buys the next book. A reader who feels cheated by a climax rarely explains it well. They say the ending was rushed, or too neat, or that they saw it coming, and then they don't click through to book two. Across a series that adds up to the only number that matters: how many readers of book one reach the end of book five.

A flat climax is almost never a problem with the climax. What you're looking at is a setup problem that only shows up at the climax, and that's why it eats so much time. Writers rewrite the scene over and over, and the scene was never at fault. The fix is nearly always earlier in the book, in the twenty or thirty pages that were supposed to buy the winning move and didn't. Once you know that, four weeks of polishing a final showdown can become an afternoon with a spreadsheet and a red pen. Revision covers the pass, and the steps below show how to run it.

When you know your climax before you write the middle, you know what the middle is for. A middle with no known destination fills up with stuff. A middle that knows what the last chapter needs becomes a shopping list of things to set up. Writers who plan often say this is the most useful thing an outline does, and writers who don't plan often say it's the thing they wish they'd known at sixty thousand words.

Climaxes fail in six ways, and each has a test. All six lead readers to say the same thing, that the ending fell flat, and each needs a different fix, so run the tests in order. On a finished draft it takes under an hour.

The first failure is the climax the protagonist watches. She's there, the scene is about her, and somebody else does the deciding thing. A sibling confesses, a lawyer produces the document, a soldier shoots the antagonist, or an institution finally acts. This is the most common fault of all, and it's almost always an accident: the writer needed the plot wrapped up and reached for whoever was handiest. The test is to write the climax as one sentence in the active voice, with a single main verb, and look at the subject. "The protagonist forces the inspector to reopen the case" passes. "The inspector reopens the case after Bracher's neighbor calls the ministry" doesn't, and no amount of Bracher's thoughts and feelings in her viewpoint will fix it. If she isn't the subject of the deciding verb, she didn't have a climax.

The second is resolution by a means that turned up two pages earlier. The winning move is there and she makes it, but the means arrived just in time: a contact she's never mentioned, a document she suddenly remembers, a talent the book has never shown her using. On paper that counts as a setup. To the reader it feels like a rescue, because they had no chance to take it in. The test is to find the page where the winning means first shows up in the manuscript and ask what else that page is doing. A good setup pays its own way. The skill was shown because it cost her something, the contact came in because she needed him for a different reason, the fact was planted in a scene about something else. A plant that shows up late and sits there waiting to be used is obvious to readers even when they can't name it. Foreshadowing is the craft of hiding the payment in plain sight.

The third is the antagonist who gives up, or turns out to have been wrong all along. The opposition melts away instead of being beaten. The villain's plan turns out not to work, or the villain has a change of heart, or the villain was working from a mistake that one conversation clears up. The test is to write one paragraph of the climax from inside the antagonist's head, using only what he wants and what he's willing to spend. If he fights as hard as his want will take him, you have opposition. If he folds, or if the paragraph is hard to write because you don't know what he wants badly enough to fight for, the problem isn't in the climax at all, and Antagonist is where to go.

The fourth is a climax that resolves the wrong question. The scene is strong, the protagonist drives it, the setup is clean, and it settles something other than what the book opened. Call it the obligatory-scene failure. Of the six it's the hardest to see from inside a draft, because every part of the scene on its own is working. The test is to write two sentences separately and compare them: the question your opening fifty pages put in the reader's head, and the question your climax answers. Look at the nouns. If the first sentence says "brother" and "prison" and the second says "wedding" and "forgiveness", you've found the fault, and the fix is in the book's structure. Dramatic Question helps you write both sentences truthfully.

The fifth is a climax that arrives too late for the reader to feel it. This one comes in two forms. In the first, the outcome stopped being in doubt long before the scene, so the reader is watching a formality. In the second, there's so little book left that the climax has no room to be felt, and it reads like a summary of an ending instead of an ending. The test is to find the last page where a reasonable reader could still have been unsure how this ends, and measure the gap between that page and the climax. A big gap means the tension died early, and the fix is in the pages between, not in the scene. Then measure the other way. If only a handful of pages separate the climactic moment from the final line, check whether you've given the reader anywhere to put what just happened.

The sixth is the false climax, followed by a smaller one. The book delivers a big confrontation and the reader braces for the ending. Then a second, quieter confrontation happens afterward, and it's supposed to be the one that counts. Sometimes that's on purpose and excellent, as in the outer-then-inner order described above. Usually it's an accident of drafting, and it reads like an ending that won't stop. The test is to list the last four set pieces and write beside each one what it cost the protagonist, leaving out what happened in it and writing only the cost. If the biggest cost isn't the last item on the list, you've found the problem. The fix is one of three things: merge the two scenes, cut the earlier one, or clearly demote it so the reader knows it's a stage and not the finish, which usually comes down to what's still unsettled when it ends.

A few genre expectations at the climax have no give at all, because they're what the reader paid for. Most "rules" of genre are habits with plenty of give in them. The five that follow you can break only as a decision, knowing you're doing it.

In a mystery, the solution is delivered. The reader of a whodunit has been invited to compete, so the climax owes them the name, the method and enough of the reasoning to check the work, and the reasoning has to rest on clues the reader was given. A mystery that holds back the solution, or names a culprit on evidence the reader never saw, has broken the only promise the genre makes. Everything else can bend. The detective can be wrong for most of the book, the culprit can escape and justice can fail, but you can't skip the verdict. See Mystery and Cozy Mystery, and Red Herring for how to play fair with clues.

In romance, the climax delivers union, or a refusal the book has argued for. Romance readers come for the couple together at the end and secure in it, and the Romance Writers of America's own definition of the genre puts an emotionally satisfying and optimistic ending at its center. A romance that splits the couple up is something other than a sad romance, and it'll be reviewed that way. The interesting case is the refusal on purpose, a novel that takes the shape of a love story and turns down the union deliberately. That can be superb, on one condition: the refusal has to be the answer to the question the book actually asked, not a swerve away from it. If the book has spent four hundred pages asking whether these two can be together, and the answer is no because being together would make one of them disappear, the reader has been told the truth. What matters is whether the book argued for the refusal or forced it in. See Romance and Romancing the Beat.

In horror, the question is survival, and the reader must see the answer. Horror's climax settles whether the thing can be survived, and either answer is open to you. William Peter Blatty's The Exorcist (1971) moves the possession into the priest himself and then has him throw himself from the window, so the child lives and the priest doesn't, and both halves come from the same act. What horror can't do is leave the question hanging. Some horror novels end with the threat neither beaten nor winning, just still around, and those books rely on the reader to supply an ending, which works in a short story and rarely at novel length. See Horror.

In a thriller, the protagonist is in the room. The rule isn't that the hero wins. The face-off is physical and direct, and she's inside it rather than hearing about it. A thriller that wraps up offstage, by an arrest or a news report, has kept its plot and thrown away its form. See Thriller and Ticking Clock.

In fantasy and science fiction, the invented system has to be the means, not the excuse. If a world has rules, the climax is where a reader checks whether you meant them. The most satisfying ending these genres have uses a power of the magic or the technology that the book already set up, at a cost the book already set up, and the least satisfying one needs a new power. See Magic System, Epic Fantasy and Science Fiction.

Writing a series gives you a real problem with structure at the climax, and there's a real way to solve it. Book one has to end. A reader who finishes it has to feel that something was settled, or they'll call it half a book, which is the most damaging review a series opener can get. You also need a reason for book two to exist. The wrong fixes are common enough to name. Ending on a cliffhanger instead of a climax settles nothing and turns the reader's satisfaction into annoyance. Answering the book's question and then tacking on a final chapter where a shadowy figure announces a bigger threat is a bolt-on, and readers feel the join. And splitting one book's climax across two volumes is the version that brings in refund requests. The fix is to run two questions at two sizes and answer only the smaller one. The book's question closes properly, with a confrontation and a verdict, while the series arc question stays open, and, most important, it's raised by the climax instead of being parked to one side. The best version is when the same act that settles book one also sharpens the series question. She wins, and winning shows her what she's really up against, or what she's now ready to do. Nothing has been held back, the reader has a complete story, and they're left with a bigger question than they had at the start. Series covers the planning side.

Real-world examples

Suzanne Collins's The Hunger Games (2008): the winning move, planted twice

Scholastic published The Hunger Games in September 2008, and it's still one of the most widely read commercial novels of the century, which makes its ending a handy shared example. What follows gives the ending away, though it's famous enough that most readers know it.

The climactic sequence runs in stages. First come the mutated creatures, then the long night on the Cornucopia and the death of Cato. Then the Gamemakers announce that the rule allowing two winners has been taken back, so Katniss and Peeta now have to kill each other. Instead, Katniss brings out the nightlock berries and the two of them get ready to eat them, which would leave the Games with no winner at all. The Gamemakers can't accept that, so they stop the Games and declare them both winners.

Look at where the berries come from. They were planted twice, and neither planting was free. Katniss knows nightlock because her father warned her about it, in a line about being dead before the berries reach your stomach, and that warning is a piece of family know-how from the life the book has spent chapters setting up, along with the foraging and the archery. The reader has also already watched the berries kill, when Foxface dies from eating the ones Peeta picked without knowing what they were. By the time Katniss reaches for them, readers know exactly what they do, because the book made them watch.

Then look at who wins. Katniss is plainly the subject of the deciding verb, and what she does isn't violence but a refusal aimed at the people watching, which is the inner climax. The book's inner question has been whether she's a survivor who does what the Capitol demands, or someone who won't, and the berries answer it with the same gesture that answers the outer plot, so outer and inner land together.

The ending also raises the series question instead of parking it. The Games are over, and Katniss has become a problem for the people who run them, which drives the next two books. Nothing was held back to get there, because holding something back would have made it a cliffhanger, and this is the opposite.

Claire Keegan's Small Things Like These (2021): the quiet climax, and the book that stops there

Faber published Claire Keegan's novella in the United Kingdom in October 2021, and Grove Press published it in the United States the same year. It was shortlisted for the Booker Prize and won the Orwell Prize for Political Fiction in 2022, and it was filmed in 2024 with Cillian Murphy. At around 128 pages you can read it in an evening, which makes it unusually easy to study. The ending is the whole point of the book, so this example gives it away.

Bill Furlong is a coal and timber merchant in New Ross in December 1985. While delivering fuel to the local convent, he finds a young woman, Sarah, shut in the coal shed. The convent is running one of Ireland's Magdalene laundries, and the whole town knows and doesn't say so. In the climax, Bill goes back, takes her out, gives her his coat and walks her through the town to his own house, and then the book ends.

Nothing explodes and nobody has a long showdown. The Mother Superior isn't defeated, the institution isn't brought down, and the reader is left in no doubt that what Bill has done will cost him and his family badly. And still it's one of the most forceful climaxes in recent fiction, for three separate reasons.

The act is entirely his. Nobody helps him and nobody nudges him, and the town has spent the novella making it clear that everyone would rather he didn't. The deciding verb is his.

The act is paid for, in a currency the book has been saving up since page one. Bill was born to an unmarried teenage servant, and a kind employer kept her on instead of sending her to an institution. He spends the novella circling that fact and what it means that his own life came down to someone else's small decision. The act at the end is that decision made again, by him, the other way around. Everything in the climax was bought earlier, and none of the buying looked like setup at the time.

And it's the point where the most is at stake and the reader is least sure, which is what "peak of tension" means. The reader doesn't know until the last pages whether he'll do it, and knows full well what doing it will cost, which is tension without any noise in it at all.

The book has one more lesson, for anyone worried about length. It has almost no resolution. The climax is the ending and the aftermath is left to the reader, which is a risky choice that works here because the costs are so specific that the reader can work them out alone. Most novels need the pages this one leaves out, and what those pages are for belongs to Resolution.

Arthur Miller's Death of a Salesman (1949): the climax that is lost, and the want that was wrong

Miller's play opened at the Morosco Theatre on Broadway on February 10, 1949. It ran for 742 performances and won the Pulitzer Prize for Drama that year. It's in print everywhere and can be staged anywhere, and it's the clearest proof there is that a lost climax can be the strongest kind. The ending is nearly eighty years old and has been on school reading lists for most of them, but an ending is still an ending, so here's your warning.

The climax comes in two parts, and they're separate scenes. First, in the house, Biff finally tells his father the truth: neither of them is a great man, and everything Willy has built his hopes on is false. Then Biff breaks down and holds him. Willy doesn't hear a word of what Biff said. He hears that his son loves him, and he decides the dream is still alive after all. Second, that same night, he drives the car into a crash so that Biff will get the life-insurance money and have the funds to become the man Willy wanted.

That ending does some things that are hard to do any other way. The deciding act is the protagonist's, and it destroys him, which is what a tragic climax needs and why tragedy isn't a passive form. Willy is more active in the last ten minutes of the play than anyone else in it.

He also gets what he wanted. This setup is the hardest to write, and it hits hardest when it works. Willy has wanted to be worth something to his son, and the money is his answer, so he dies believing he's delivered it. The Requiem is where the play shows how false that want was. Biff turns down the inheritance and the life it was meant to buy, and Linda can't understand why nobody has come. Willy had pictured a funeral crowded with buyers from every state, and only his family and two neighbors come.

And the inner climax comes before the outer one, which makes the ending unbearable instead of merely sad. Biff's breakdown is the moment of truth, and Willy has every chance to see it. He misreads it, and then he acts. All the way through that final act the reader or audience knows something the protagonist doesn't, which is dramatic irony doing work a car chase can't. Tension explains why it's the most reliable engine there is.

Anthony Horowitz's Magpie Murders (2016): the obligatory scene, withheld on purpose

Orion published Magpie Murders in the United Kingdom in October 2016, and HarperCollins published it in the United States in 2017. It was adapted for television in 2022. The book is a mystery built inside another mystery, and it's on this page because its design shows the obligatory scene in action instead of arguing about it.

The book gives you a complete golden-age detective novel, an Atticus Pünd mystery by the made-up novelist Alan Conway, set in an English village in the 1950s. You read it the way you'd read any whodunit, with suspects, clues, a country house and a detective piecing the pattern together. Then, just as Pünd is about to name the murderer, the manuscript stops, because the final chapters are missing. Susan Ryeland, the editor holding the pages, can't ask the author where they went, because he's dead. Telling you this spoils nothing, since it's the premise.

What makes the book useful is the feeling it gives you, because it's a controlled test of what a missing climax does to a reader. Every convention has been honored. The prose is good, the clues are in place, the detective is charming and the village is fully built. Losing one scene wipes all of it out. The reader isn't mildly curious. The reader is owed. That reaction is just what the French critics of the 1800s were describing, and the novel's outer plot then runs on it: Susan's hunt for the solution is a hunt for the obligatory scene, and the reader wants her to find it as badly as she does.

For a writer the lesson is how little it takes, and copying the trick would miss it. Take out one scene, and a solid, complete novel becomes unbearable. Your third act is carrying that much weight.

William Archer's Play-Making (1912): where "the obligatory scene" comes from, and what is actually established

Craft writing uses the term all the time, usually with a confident claim about who came up with it. The original source is free to read, and it says something more careful than those claims do.

William Archer (1856 to 1924) was the Scottish critic and translator who introduced Ibsen to English readers, and he published Play-Making: A Manual of Craftsmanship in 1912. Chapter XIII is titled "The Obligatory Scene", and it opens by refusing to settle who coined the term. Archer writes that he doesn't know whether Francisque Sarcey invented the phrase scène à faire, but that it surely owes its wide use to him. Sarcey (1827 to 1899) was the drama critic of Le Temps, and Archer notes that he used the phrase so often, in so many settings, that it can't be pinned to a strict definition.

So here's what the record shows, and what it doesn't. It shows that French critics were using the phrase in the later 1800s, and that Sarcey is the critic who made it popular. It shows that Archer carried it into English as "the obligatory scene" and gave it a working analysis. What it doesn't show is that Sarcey invented it, and the man who brought it into English said so in the sentence that introduced it. Anyone who tells you Sarcey coined the term is going further than the source does.

The useful part is what Archer did with it. In place of Sarcey's loose usage he offers a definition: a scene the audience foresees and desires, and whose absence it may with reason resent. Then he sorts the reasons a scene becomes a must into five kinds. A scene can be made obligatory by the logic of the theme, or by the demands of dramatic effect. It can be made obligatory by the author's own apparent build-up toward it, or by the need to justify a change of character or will that's too important to take for granted. And it can be made obligatory by history or legend.

Three of those five carry straight over to novels, and the third is the one to tape to the wall. The author himself may have rendered a scene obligatory by seeming unmistakably to lead up to it. Your genre's promises aren't the only ones that bind you. You're bound by every promise your own draft made along the way, including the ones you made by accident and forgot, which is why a reader who can't say what's wrong with your ending can still be sure something is.

Archer's chapter XVIII, "Climax and Anticlimax", is worth the extra half hour, because it explains why last acts are harder than first acts. Crises in real life have clear beginnings but almost never clear ends, because they fade out instead of finishing, so the playwright's hardest job is finding a crisis with an ending that feels both true and powerful. If you've rewritten a final chapter five times, you've met that problem without having a name for it.

Martha Wells's All Systems Red (2017): the book closed, the series left open

Tor.com published All Systems Red, the first of the Murderbot Diaries, in May 2017. It won the Nebula Award for Best Novella for 2017 and the Hugo for Best Novella in 2018, and the series has since grown to a shelf of books and a television adaptation. You can read the novella in one sitting, which makes it a quick way to study a series opener's climax. The ending is described below.

A security unit has hacked the governor module that controls it, and it calls itself Murderbot. The unit is hired out to a team of scientists doing a survey. Another team on the planet is found murdered, and the attacks turn out to be the work of a third, secret expedition, GrayCris, covering up what it found. The climax is a plan. Murderbot opens talks with GrayCris purely as a distraction, so the scientists can get into the habitat's systems and fire an emergency beacon. It works, though Murderbot is badly damaged shielding Dr. Mensah from the explosion of the launch.

Notice that the plan is the protagonist's, and that it works because of the one thing the novella has set up about Murderbot from its first page. It knows just what a SecUnit looks like to humans, and it can use that. The winning move comes straight out of what the book is about.

Then notice where the inner climax sits. After the plot is resolved, Mensah buys Murderbot's contract and offers to take it home to Preservation, where it could live as a person, and it slips away on a cargo ship instead. That's a separate, quiet scene, and it's the true ending. The outer question is who's killing survey teams and whether these people can survive, and it closes with a verdict. The inner question is what a thing like this wants when nobody is making it do anything, and the book answers that only far enough to sharpen it.

A series opener wants that shape. The book is complete, nothing was held back, and the climax created the question the next book runs on instead of having it bolted on afterward.

An everyday scenario: the climax that happened to somebody else

Here's the fault in the form most writers actually run into. Say you've drafted a novel about an orchard grower, Niamh Bracher, whose trees are inside a blight quarantine zone and set to be destroyed. The draft is ninety thousand words long and the middle works. Every attempt to get the order reviewed fails and costs her something, and by chapter thirty-one she's lost her contracts, her brother's goodwill and most of her money.

In your draft, the climax is a meeting at the ministry office, and the order to destroy the trees is reversed. Your beta readers all say the ending is flat, and two of them say they cried at chapter twenty-six and felt nothing in chapter thirty-four. That's useful to hear, and confusing too, since chapter thirty-four is where the protagonist wins.

Run the six tests, and test one finds it in a minute. Write the climax as one active sentence: The inspector reverses the order after Niamh's neighbor, Hesper Ludd, calls the ministry about the sampling error. Niamh isn't the subject. She's in the room, she's the viewpoint character and she has feelings about all of it, and she isn't the reason anything happens. Hesper is the protagonist of your ending.

Test two then finds the other half of the problem. The sampling error is what actually wins, and it first appears on the page four pages before the meeting, in a paragraph whose only job is to bring it in.

Now for the fix, which is earlier in the book and easier than it looks. In chapter nine, Niamh is grafting old rootstock and arguing with an inspector about which trees should be saved. That scene already exists, already costs her something, and already shows that she knows more about these trees than the people with the clipboards. You don't need a new chapter. You need the sampling error to be something she notices, because of what she knows, in a scene where she's looking for something else, and you need her to be the one who forces the meeting and makes the case in it. Hesper keeps her phone call, and now it's what gets Niamh through the door rather than what wins.

Nothing in the plot has changed. The order is still reversed, at the same point, in the same room. What's changed is who caused it, and that's the difference between a chapter thirty-four that makes people cry and one that reads like a summary of a good outcome.

Try it yourself

This takes an afternoon on a finished draft and about twenty minutes on an outline. Have the manuscript open and something to write on.

  1. Write the climax as one active sentence. Use one main verb (subject, verb, object), and leave out extra clauses and adverbs. Then look at the subject. If it isn't your protagonist, stop here, because you've found your problem, and everything below comes second.
  2. Name the winning move. In one phrase, write what actually settles it: the knowledge, the skill, the ally, the object or the refusal. Be specific. "Her courage" isn't a winning move. "She tells the committee what she saw in the shed" is.
  3. Find the page where that move was bought. Search the manuscript for the earliest place the winning move shows up, and write down the page number and how far through the book it is, as a percentage. Then write one sentence on what else that page is doing. If the truthful answer is "introducing the winning move", you have a plant instead of a setup, and step 6 is your fix.
  4. Write the two questions and compare the nouns. On one line, write the question your first fifty pages put in the reader's head. On the next line, write the question your climax answers. Circle the nouns in both. If the nouns match, you're answering the right question. If they don't, the ending itself is fine, and the book is about something other than what you thought. If step 4 fails, don't go on to the later steps. Read Dramatic Question and decide which of the two books you're writing.
  5. Separate the inner climax from the outer one. Write down the confrontation that settles the plot and, separately, the choice that settles the arc, and give each a page number. You'll find one of three things. They might be the same scene, which is good, so stop there. They might be different scenes in a sensible order with not much between them, which is also fine, so check the gap. Or one of them might not exist, which is your flat ending, and the missing one is almost always the inner one.
  6. Run the setup-hunting pass. This is the fix, and it's the most valuable hour on this page. Take the winning move and ask where else in your current draft it could have been bought. Don't look for places to add a scene. Look for a scene you already have that could pay for it. Find scenes in the first half where the protagonist is doing something close to it. She might be using the skill for a different purpose, meeting the ally over something else, getting the fact wrong or refusing to look at the weakness. Then rewrite those scenes so the payment happens inside them, while they're busy with something else. Two or three sentences in each is usually enough. This works because readers accept setup they weren't asked to notice. What they reject is setup that announced itself.
  7. Rank the last four set pieces by cost. For each of your final four big scenes, write what it costs the protagonist, skipping what happens in it and writing only the cost. If the biggest cost isn't the last one, you have a false climax. Merge the scenes, cut one, or leave something important unsettled at the end of the earlier scene so the reader knows it's a stage.
  8. Find the page where the outcome stopped being in doubt. Read the last third with one question in mind. At what point could a reasonable first-time reader have felt sure how this ends? Note the page. The distance between that page and your climax is the length of your flat stretch, and the fix is in that gap, not in the climax.
  9. Write the losing version. In one paragraph, write how this scene goes if she fails. If you can't write a version the book could believably have taken, the reader can't feel one either, and then no amount of rewriting the winning version will create tension. See Stakes.
  10. Check the genre obligation. Write one line naming the scene a reader of your category came for, then make sure your climax is that scene. If you're refusing it on purpose, write one more line saying what you're giving them instead, and why it answers the same question.

Climax, one active sentence: __________ Winning move: __________ Bought on page __ (__% through), in a scene whose other business is __________ Question the opening asked: __________ Question the climax answers: __________ Outer climax, page __ . Inner climax, page __ . Losing version: __________ Genre obligation: __________

Common mistakes and misunderstandings

Loudest is not highest. The most common misreading of "peak of tension" is that the climax should be the biggest event in the book. It should be the point where the most is at stake and the doubt is greatest, and a signature on a document can beat a burning building. If your ending is louder than your midpoint but no less certain, you've turned up the volume and left the tension where it was.

The fault is almost never in the scene. A flat climax is a setup problem that only shows up at the climax. Writers who don't know that can spend a month on forty pages and improve nothing, because every version of the scene has the same unpaid-for winning move inside it. Work out what's wrong before you polish, and use step 6 above for the fix.

A climax is not a revelation. Information changes what the reader knows. Action settles what happens. A plot twist at the end can be superb, and on its own it isn't an ending, which is why some very clever books get reviews that praise how they work and never mention a feeling. Ask what your protagonist does on the last page of the confrontation.

A climax does not have to be won. Half the endings open to you sit outside victory: tragedy, the outer loss that's an inner win, and the want that's granted and exposed. Readers need a verdict the protagonist delivered, even if it doesn't go her way. If an ending won't work, write the losing version before you add another obstacle.

The character who explains her arc. A protagonist who says what she's learned has just given a report. The reader needs to watch the change cost her something as it happens. If the only sign of the arc in your climax is dialogue about the arc, the inner climax is missing, and you've written a summary of it.

Letting the setup announce itself. This is the opposite mistake to the two-page rescue, and it does nearly as much damage. If a gun is described lovingly in chapter three and does nothing else in the chapter, the reader hears that as a promise. Archer's formula is what you're aiming for: foreshadow without forestalling. Bury the payment in a scene that's about something else. Foreshadowing has the technique.

The low point and the climax are two scenes, not one. They sit next to each other and do completely different jobs. In one, the protagonist is at her worst and stripped of her means. In the other, she acts from whatever she has left. Drafts that merge them read like an ending that arrives, goes flat and then arrives again. All Is Lost owns the first.

A series opener that withholds instead of closing. An unanswered question at the end of book one isn't a hook, just a debt. Readers buy book two because book one satisfied them, and being irritated by it won't get them there. Close the book's own question, and let the climax raise the bigger one.

Aftermath placed before the final beat. Once readers know the answer, their attention shuts down, so any explanation, reunion or reflection that comes before the last beat of the confrontation will feel like the book stalling. Put all of it after, and keep it short. That part belongs to Resolution.

The third act executes; it does not decide. By the time you reach the climax, the choices that matter were made two hundred pages earlier, in the setup. Some writers reach 80 percent with no idea how the book ends, and they usually aren't short of ideas for an ending. They're finding out that the first half didn't buy them one.

Related elements