Back to the table

48 Intermediate Ch Chemistry Character

Chemistry

The pull between two characters that readers can feel.

What it is

Chemistry is the reader's sense that two characters are charged when they're together. A scene with both of them in it feels more alive than a scene with either of them alone.

Notice where that sense lives. It lives in the reader, not in the characters. That may sound like a quibble, and it's the most useful idea on this page, because it turns a vague hope into a practical question. Nobody can answer "Do these two have chemistry?" You can answer "What on the page produced that feeling in the reader, and could I do more of it?" in an afternoon.

Writers get stuck here because of the word. In the everyday sense, chemistry belongs to two substances. Put them in a jar and something happens. Characters aren't substances. They don't react on their own, because they don't exist until you write a sentence, and no sentence can put the effect there directly. They had chemistry produces nothing. She had never felt anything like it produces nothing either. The air between them seemed to thicken produces, at best, weather.

The reader always works out the pull from evidence, the same way they work out that a character is frightened from a hand that won't stay still. Show, don't tell is the general rule, and chemistry is where writers break it most often and notice it least. From inside a viewpoint character's head, a stated feeling looks like a fact.

This page is narrower and more hands-on than the one on who the other person needs to be. Given two characters, how do you make the reader feel the charge between them, line by line? Love Interest covers the person: a full human with their own life instead of a trophy, and whether they're a co-protagonist or a subplot.

Chemistry usually gets described with a list: contrast, banter, vulnerability, and an obstacle. The list is right, and on its own it's useless, because a list invites you to tick items off. What helps is knowing what each one is for.

If you take one working method from this page, take this one. Chemistry is mostly built out of what each character notices about the other, and what they do once they've noticed.

Attention costs less than any other material in fiction, and it persuades more. It costs no plot, hardly any words, and no shuffling of your outline, and it works in one viewpoint or two. A reader almost can't disbelieve it, because nobody chooses what they notice, and we all know it. It comes in three layers, and you want all three running:

Both people noticing matters more than how hard they notice. One character noticing everything about another who notices nothing gives you longing instead of chemistry, and longing is a different effect, often a sadder one. If you want a charge, both people have to be paying attention, even if only one of them has a viewpoint. In single-viewpoint fiction the reader can't get inside the other person's head, so their attention has to show in what they do: showing up with the right thing, asking the question nobody else asked, remembering the detail.

What makes a pair interesting is that they want slightly different things out of the same meeting, and writers use this trick less than any other in the element. Most flat romantic scenes are flat because both characters want the same thing from the encounter. Usually it's "to be near each other", and that leaves nothing to watch.

Give them goals that sit side by side and don't quite match, and the scene gets an engine without needing a plot. She wants to borrow a tool and get out of the workshop without a conversation. He wants to be asked for help, because being asked is the only way he knows of being wanted. Neither of them says any of that, and both of them are, technically, having a conversation about a torque wrench.

"Is the three-eighths in here?" "Bottom drawer." Emlyn did not get up. Thea found it, weighed it, put it back, and took the smaller one instead. "This'll do." "That's the wrong size for that bracket." "It's the right size for what I'm doing to the bracket." "Right." He turned a page. "Well. Shout if the bracket wins." She got as far as the door. "You could come and look at it." "I could," Emlyn said, and put the magazine down before she had finished asking.

Nothing has happened, yet both of them got something, and neither of them said what it was. The scene has chemistry in it, and it's made of nothing but two agendas and a wrench. The formal name for the machinery is a scene goal for each character, and the move here is to make them different goals that the same room happens to serve. Asymmetry also gives you your escalation for free. As the book goes on, the gap between what each one wants from an encounter narrows, and the reader can feel it narrowing. You get a character arc you can measure.

Now for the part that goes against instinct, and it's where most manuscripts lose the effect they were aiming for. The pull a reader feels comes from the gap between wanting and having, more than from how much the two characters want each other. Attraction is the pressure and the barrier is the valve. Open the valve and the pressure disappears, however high it was.

So you usually answer the question "how do I give these two more chemistry?" by working on the obstacle instead of the attraction. If a scene between them feels dead, the fault is rarely that they don't want each other enough. Nine times out of ten, nothing is stopping them, so the wanting has nowhere to go.

A barrier holds when it comes from who these people are, and when neither of them can dissolve it by being sensible for ninety seconds. A professional conflict of interest, a clashing obligation, a belief about what being loved costs, a promise already made to somebody else: these survive contact with adults. A half-overheard remark doesn't. The moment a reader suspects the obstacle is fake, they stop worrying, and tension is made of worry. Love Interest has the structural version of this argument, including what romance readers named the Big Misunderstanding.

The flip side is something nobody warns you about. Removing the obstacle usually kills the chemistry, and it kills it on the page where you remove it. That doesn't mean you have to keep two people apart forever, because that creates its own ugly effect, where readers feel manipulated and start skipping. What you do need is to know in advance what the pair will run on afterward, because you'll need a second engine. The usual answers are a shared outside problem, or a new and smaller disagreement that comes from being together instead of getting together. Or it can be an internal conflict that the relationship has now made worse instead of better. Decide this before you write the scene where they finally get what they want, not after your beta readers tell you the last third went quiet.

Physical awareness is the part writers most want a formula for, and the formula is dull: be specific, be brief, and stay in one body. Generic awareness reads as filler, because it could be any two people. Her heart hammered. She was intensely aware of him standing behind her. Electricity shot up her arm. Those lines are only labels for observations, and the reader has seen them a thousand times. The version that works names something odd and small:

He leaned past her to reach the switch, and she held still, the way you hold still for a wasp.

The sentence has nothing explicit in it, and it tells you her body has an opinion she hasn't agreed to. It would work in a cozy mystery, in a young-adult novel and in an erotic romance, because specificity is portable and heat isn't. Build the awareness out of concrete, slightly unexpected physical fact, and you can set the explicitness anywhere afterward without rewriting the chemistry.

That matters because the range is huge. Sweet and inspirational romance lines carry a closed-door promise, and their readers will complain in reviews if it's broken. Contemporary romance sits almost anywhere on the scale, often signaled on the cover and in the blurb. Romantasy readers often expect explicit scenes and say so. Crime and literary fiction usually want awareness without sex on the page. These are reader promise questions, and they have nothing to do with craft quality. A writer who mixes the two up ends up believing that more explicit means more chemistry, and it doesn't. Explicit scenes with no awareness in them read as choreography, which is the most common complaint about them. The most charged scene in a book is very often one where two people are fully clothed and three feet apart.

A list of physical and behavioral signals helps here, and Ackerman and Puglisi's The Emotion Thesaurus is the standard one. You don't copy from it. The second or third signal you reach for on your own is always the cliché, and a list gets you past it.

This element sits in the character row instead of the romance row for a reason. Every mechanism above works on pairs who will never kiss, and a general craft table earns its keep there.

"You've redone the kitchen." "Last spring." "It's the tap I noticed. My wife wanted that tap." Mensah turned it on, and off. "Eleven hundred pounds, that tap." "Nine hundred." "Nine hundred," he said, and wrote it down slowly, so that she could watch him do it.

The exchange has contrast, attention, asymmetry, a barrier, and a small unguarded moment where she corrects him because she can't stand to be thought extravagant. There's no romance anywhere in it, and it's the same machine.

Why it matters

Chemistry is what readers name when they buy an author again. Reviews and recommendations for series fiction are full of it. People say they came back for the pairing, and they mean the two characters together, not either one alone. A reader uses up a plot once. A pairing can run for twenty books, so treat it as an asset and not as atmosphere.

It's the least costly upgrade you can make in revision. Every other fix for a flat relationship costs structure. Adding a subplot costs chapters, and changing the obstacle costs a rewrite. Adding attention costs a sentence a scene, and you can do it in one pass with a marker. For a writer turning out several books a year, that trade is the whole argument.

A strong pairing carries a slow middle. In the part of a book where the plot has to be in transit, two charged people talking is enough to keep a reader turning pages. Books with a strong pairing survive a sagging middle, and books without one don't. The sign is a reader who puts the book down at forty percent, meaning to come back.

In romance, chemistry is what separates a competent book from one that sells. The relationship is the plot there, so the felt pull is the product, and nobody reads it as decoration. That's why a structurally perfect romance with the right beats and no charge gets three-star reviews saying the prose was lovely and something was missing. No other element has this wide a gap between technically correct and actually working.

Outside romance, a charged pair is the most reliable way to make a two-hander readable. Say your book has a detective and a partner, a heist crew with two leads, or a mentor who's also a rival. The pairing is doing far more work than you probably think, and handily, it's also the easiest thing to strengthen without disturbing the plot.

When a manuscript is "fine", missing chemistry is the most common thing a developmental editor points to. A developmental editor is the editor who looks at the big picture. You get "fine" when two people in a room have no contrast, no barrier and nothing to notice. The note usually comes as a feeling instead of a diagnosis (I didn't mind whether they got together), and this page is the list of places to look.

Real-world examples

Shakespeare's Much Ado About Nothing (about 1598): banter with a grievance under it

Much Ado About Nothing is usually dated to 1598 or 1599, and it was printed in quarto in 1600. Beatrice and Benedick are the oldest surviving model for the pair whose banter is the plot. You can see why they still work in the first scene, which takes about ten minutes to read for free on the Folger Shakespeare Library's site. The jokes aren't what makes them work.

To start with, the play tells you right away that the wit is a condition, not a personality. Leonato calls it "a kind of merry war" between them, which sets up what follows as an ongoing battle instead of two clever people meeting.

Then Beatrice's first exchange with Benedick ends with her saying she knows him of old. It's a throwaway line, and it plants a shared history the audience never gets to see. Something happened between these two before the play started, and the play never fully explains it. Leaving it unexplained is the point, because every barb afterward lands on a bruise the audience can sense but not see. Both of them have sworn in public never to marry, so a plain statement of feeling is off the table. They can't simply say it, so they say everything else.

Watch, too, what the play does when it needs to turn banter into something with a cost. (This gives away a mid-play turn, so skip the rest of this paragraph if you don't know it.) After Hero is shamed in public at her own wedding on a made-up charge, Beatrice and Benedick are finally alone. They get as far as admitting they love each other, and then Beatrice asks him to kill his best friend. They're the same two people, talking the same way, and it has stopped being funny. The vulnerability arrives, the barrier is instantly replaced by a much worse one, and a comedy grows teeth. There's the whole mechanism of this page, carried out four hundred years before anyone wrote it down.

For a working writer, the useful takeaway is the bruise. If your banter feels like a performance, the missing ingredient is usually history. Give the two of them something that happened before page one that neither will name.

Moonlighting (1985 to 1989): the obstacle removed, and the curse named after it

Moonlighting, created by Glenn Gordon Caron for ABC, premiered on March 3, 1985 and ran until May 14, 1989. Cybill Shepherd played Maddie Hayes and Bruce Willis played David Addison, and together they ran a Los Angeles detective agency. The show's fame rests almost wholly on the pairing: two people with a complete conflict of method, arguing at speed, held apart by a working relationship neither of them wanted.

In the episode "I Am Curious… Maddie", broadcast on March 31, 1987 as the second-to-last episode of the third season, the two finally sleep together. It was the most-watched episode the series ever made, and over the next two seasons the show fell apart. It had finished the 1986 to 1987 season ranked ninth in the Nielsen season averages and twelfth in 1987 to 1988. By 1988 to 1989 it was outside the top thirty. Television writers gave the pattern a name, the "Moonlighting curse", and ever since, people have used it as a reason not to let a will-they-won't-they couple settle things.

Be careful with it, though, because the cause is disputed, and the dispute teaches you something. Scott Ryan interviewed the cast and creator for an oral history of the series, and he argues the decline came from a production collapse instead of from the romance. Shepherd's pregnancy with twins, and Willis's film schedule and injury, meant that much of the next season was made with the two leads barely appearing together. Episodes were delayed and pre-empted, too. On that account, the damage came from separating the pair, and uniting them had little to do with it.

Both readings point the same way for a novelist, which is why the example earns its place. Whether the charge died because the barrier came down or because the two stopped sharing scenes, the audience lost the thing this page is about. If you're writing the moment your pair gets what it wants, don't ask whether to write it. Ask what they'll be doing together on the page right afterward, and have the answer ready in advance.

Dashiell Hammett's The Thin Man (1934): chemistry that survives the wedding

The counter-case is ninety years old. The Thin Man, Dashiell Hammett's last novel, appeared in a condensed version in Redbook in December 1933, and in book form from Knopf the next month. Its detectives, Nick and Nora Charles, are already married when it starts. They're comfortably off and completely sure of each other, so every obstacle this page recommends is gone before page one. It works anyway, and you can copy the reasons.

The pairing runs on asymmetry instead of suspense. Nora wants the case and finds it thrilling. Nick wants a drink and to be left alone, and he gets dragged in. That difference creates a conflict in nearly every scene they share. The conflict is affectionate, not hostile, and that turns out to be enough.

The pairing also runs on attention. They read each other's small signals faster than anyone else in the book, and they act on them without explaining, which is what married competence looks like from outside. And it runs on a barrier that's outside them and never runs out (the murder, the police, other people's lies), so the pair always has something to team up against.

MGM bought the film rights eight days after publication and made the 1934 picture with William Powell and Myrna Loy, which had five sequels. Indie authors should note the business lesson. A couple who are already together can carry a long series, as long as you swap the getting-together engine for a working one. For Nick and Nora, the new engine is a shared problem plus a permanent, low-level disagreement about how to live.

Ali Hazelwood's The Love Hypothesis (2021): one viewpoint, and attention as the whole engine

The Love Hypothesis was published by Berkley on September 14, 2021, and became a bestseller largely through word of mouth on social media. It began life as fan fiction. Hazelwood has described writing it for a fandom exchange, after a friend asked for a fake-dating story. Later an agent took it on, and she reworked it as an original novel.

The book is useful to study for a couple of craft reasons, and the main one is that it's written in a single viewpoint. Every bit of evidence for the second lead's feelings has to arrive as behavior. He's unreadable on purpose, and what the reader gets instead is a growing pile of things he does. He shows up where he doesn't need to be, remembers a detail from an early conversation, adjusts something so she isn't put out, and says nothing while doing something.

The heroine keeps misreading these as indifference or irritation. The reader doesn't, and that creates a very pleasant kind of tension made only out of noticing. It's dramatic irony applied to attraction, where the reader knows something the character doesn't. If you're stuck in a single viewpoint and wondering how to show a pull that runs both ways, read the first third with a pen.

The barrier is the other thing to study. It's professional and structural instead of emotional: an academic power imbalance that both characters can see and neither can wish away. It survives adults behaving sensibly, which is the whole requirement, and it means the tension doesn't depend on either of them being dense.

Read the novel critically as well as with admiration. A frequent criticism is that the heroine's misreadings last a beat or two longer than they plausibly should. That's the standing risk of running this much of your chemistry on one character failing to read the other. If you use the technique, the fix is to let your protagonist read a signal correctly now and then, and be frightened by it.

Chemistry with no romance in it: Amadeus (1979), Columbo (1971 onward) and Good Omens (1990)

Here are three pairings with no romance in them at all, running the same machinery.

Peter Shaffer's play Amadeus opened at the National Theatre in London on November 2, 1979. Peter Hall directed it, with Paul Scofield as Salieri and Simon Callow as Mozart, and it won the Tony Award for Best Play in 1981. Shaffer also wrote the screenplay for Miloš Forman's 1984 film, and he described the play as a fantasia on a real theme instead of a biography. To be clear, the historical Salieri didn't do what the play has him do. He's a character, not a verdict.

What the play shows is rivalry chemistry at full power. Salieri pays attention to Mozart with an intensity no lover in the piece can match. He alone can hear just how good the music is, and that's the injury. The contrast is total, and the asymmetry is clean: Salieri wants recognition, and Mozart wants a patron and an audience. The barrier is that Salieri can never say any of it out loud. The pull the audience feels between them is huge, and it's hostile all the way through.

Columbo, created by Richard Levinson and William Link, ran on NBC from 1971 to 1978 and came back on ABC from 1989 to 2003. It's built on the inverted mystery, where the audience sees the murder first, so there's no whodunit. What's left is a two-hander between the detective and the murderer, and the series is the best study you'll find of interrogator-and-suspect chemistry. Columbo's whole method is attention he makes a show of, noticing the wrong detail out loud. The suspect's whole problem is that they can't ask him to leave without confirming what he suspects. You get two agendas, one room and a perfect barrier, and the format ran for three decades on it.

Terry Pratchett and Neil Gaiman's Good Omens, published in 1990, gives you the partner version. An angel and a demon have worked opposite sides of the same street for six thousand years. They have a conflict of method, an outside problem that never runs out, and a shorthand the reader gets to half follow. Their barrier is simply who they each work for. Readers talk about that pair the way romance readers talk about couples, and that's the evidence that the effect isn't tied to one genre.

Jayne Ann Krentz's Dangerous Men and Adventurous Women (1992): what the writers themselves said the pull was made of

In 1992 the University of Pennsylvania Press published a collection of essays by working romance novelists, edited by Jayne Ann Krentz. In it, they set out to explain their own form to an academic world that had mostly dismissed it. Contributors included Krentz, Laura Kinsale, Linda Barlow, Susan Elizabeth Phillips, Sandra Brown, Mary Jo Putney, Elizabeth Lowell, Daphne Clair and Cathie Linz.

The book matters for this element because its recurring argument is structural instead of sentimental. The essays treat the pull between the leads as something the form builds on purpose, instead of something the characters just happen to have. They argue that the hero works as a real obstacle and threat instead of an ally, and that the genre's stock descriptive language works as a shared code between writer and reader. Kinsale, writing on point of view, claims that the reader doesn't simply stand in the heroine's shoes and identify outward from there. Whether you agree with any one essay doesn't matter much. What's useful is the habit of mind. These are writers asking what produced that effect, which is the question this whole page is built around.

Read it with its date in mind. The examples come from the market of the late 1980s, and the genre itself has since dropped some of the conventions being defended. Even so, the analysis of where the charge comes from has aged much better than the examples it's drawn from. Love Interest covers what changed and why.

An everyday scenario: the two characters everyone said were flat

Here's how this usually shows up in the ordinary run of drafting.

Say you have a finished thriller, and your lead, Rosalind, spends the book with a colleague called Hallam. Three beta readers have each, on their own, said some version of I didn't really feel anything between them. You go back and find, to your annoyance, that you did write the attraction. It's right there. She was more aware of him than she wanted to be, there was an ease between them, and there was something in his manner she couldn't place. There's your finding. Every one of those lines is the narration asserting the effect.

So you run the list. Contrast: they agree about everything, because you needed them to be a working team. Barrier: nothing, apart from being busy. Attention: he has three appearance notes in the whole manuscript, and she has none. Asymmetry: in every shared scene they both want the same thing, which is to solve the case.

The repair takes a weekend and touches no plot. You give Hallam a method Rosalind thinks is reckless, so they can argue. You give her one thing she keeps noticing that nobody else would, the way he answers his phone as if it's bad news. You give her one thing she does with it: she stops calling him and starts showing up. You give him one moment out of costume, thirty words long, and you give her a reaction to it that she covers. Then you give them a barrier that's real, because she's the one who'll have to write the report on him.

Word count added: maybe eleven hundred, spread across nine chapters. The next reader says she stayed up to find out what happened with those two.

Try it yourself

This works on a book you're drafting, one in revision, or a pairing you're planning. Allow about ninety minutes for the full pass, or twenty for steps 1 to 3.

  1. Strike out every assertion. Search your manuscript for the sentences that name the pull: drawn to, aware of, something about, never felt, electric, charged, chemistry. Highlight every one, and don't delete them yet. Count them first. The number shows how much of your chemistry is being claimed instead of shown, and most writers are startled by it.
  2. Run the removal test. Take one scene, cut every highlighted sentence from it, and read what's left. If nothing in the remaining text would tell a stranger these two are charged, you've found the actual problem, and steps 4 to 7 are the fix.
  3. Fill in the seven-line card. For your pair, one line each, using the card below. Any line you can't fill in is a scene you haven't written yet.
  4. Give them different agendas in the next shared scene. Write down what each one wants out of this encounter, not out of the book. If the two answers are the same, change one. The easiest change is to give the less active character a want about the relationship that they won't state, such as being asked, being refused, or being believed.
  5. Write the noticing pass. Go through every scene the two share and add just one thing one of them notices about the other, one per scene and no more. Make it specific to the observer's history or trade, and make it something nobody else in the book would pick up on. Resist the urge to make it flattering.
  6. Write the doing pass. For half of those, add a small action that follows from the noticing and is never explained. Nobody comments on it, so the reader does the work.
  7. Audit the barrier out loud. Say the obstacle as a single sentence, then ask whether one frank five-minute conversation would end it. If it would, you have a delay instead of a barrier. Replace it with something that comes out of who these people are: an obligation, a professional conflict, a promise already made, or a belief about themselves that would take the whole book to shift.
  8. Do the banter test on your best exchange. Take the funniest dialogue in the book. In the margin, write what each person is trying to get in each line. If your answers are all "to be funnier than the other one", rewrite it so that one of them is trying to find something out and the other is trying not to tell them. Keep the jokes. See Dialogue and Character Voice.
  9. Plan the second engine. Write one paragraph about what these two are like together after they get what they want. Name the shared problem, or the new disagreement, or the way being together has made one of their private troubles worse. Do this before you write the scene where the barrier comes down.
  10. Check the range against your promise. Decide how explicit the physical awareness needs to be for the readers you're writing for. Then make sure your specific details would still work one notch up and one notch down. If they wouldn't, they're heat instead of observation, and heat doesn't travel.

The seven-line card, to paste into your notes:

The pair: __________ and __________ Conflict of method (they disagree about how, not about what): __________ What A notices that nobody else would: __________ What B does that shows they are paying attention: __________ What each one cannot say, and why not: __________ The barrier, stated as a sentence a reasonable adult could not talk their way out of: __________ What they will run on after the barrier comes down: __________

Common mistakes and misunderstandings

Chemistry asserted in narration. By far the most common fault, and the hardest to catch in your own draft. From inside a viewpoint character's head, she was extremely aware of him reads like a report of fact. What it really does is tell the reader what to feel. Strike it and see whether anything's left.

Banter that is only two people being witty at each other. Wit is a texture. Chemistry is a negotiation. The test is whether each line is trying to get something the speaker can't ask for outright. If both characters are simply winning, you've written a double act, and a reader will admire it without caring about it.

Instant attraction with nothing under it. Two people meet, both are instantly and mysteriously transfixed, and the book runs on the strength of that. Attraction at first sight is perfectly usable, and it's one of the oldest openings in the form. It's a starting condition, and it proves nothing. What makes the reader believe it is whatever arrives in the next twenty pages: a contrast, a history, a specific thing noticed. Without those, the pair reads as two people obeying the outline.

The obstacle that is a misunderstanding rather than a reason. A barrier that one plain sentence would dissolve teaches the reader that nothing is really at stake. A reader who has learned that stops worrying. Misunderstandings only hold up when the characters' own natures make the silence unavoidable. Love Interest covers the structural version, including the beat-sheet habit of putting in a rupture because the template has a low point on that page.

More explicit is assumed to mean more charged. It doesn't, and the mix-up leads to the most common complaint about explicit scenes: they feel like choreography. Awareness is what makes a physical scene work. Without it, detail is just detail. In a great many novels, the most charged page has two people three feet apart and fully dressed.

Contrast that lives entirely on the surface. Tidy against messy, fancy against rough, morning person against night owl. Each is good for one exchange. Lasting contrast is a disagreement about how the world works, which sparks a fresh argument in every new situation you put them in.

One-sided attention mistaken for a pull. If A notices everything about B and B notices nothing, you've written longing. Longing is a real and useful effect, and a different one. A charge needs both people paying attention, whether or not both have a viewpoint.

Removing the barrier without replacing the engine. Writers spend a book building pressure, release it, and then find the relationship has gone quiet. It happens in the last third of a standalone, or in book four of a series. The relationship has to keep creating obligations and disagreements once it has stopped creating suspense.

Keeping two people apart past the point of belief. The opposite error, and it reads as manipulation. Once the only thing keeping your pair apart is the author's reluctance, readers feel it, get irritated and start skimming. A barrier has a natural lifespan. Plan the replacement instead of stretching the original.

The idea that this is a romance skill. Rivals, partners, siblings, an interrogator and a suspect: every mechanism here works on pairs who'll never be lovers. A two-hander in any genre lives or dies on it. If you write crime or fantasy and have been skipping the romance shelf, you've been skipping the best teaching around on your own detective pair.

Signals that repeat across a backlist. The held breath, the hand that almost touches, the sentence trailing off. Every writer has three or four defaults. Readers notice a house style in the awareness long before they notice it anywhere else, so keep a note of what you used in your last book.

Related elements