How to Write a Crime Screenplay: From Heists to Noir

Crime is the genre a reader opens expecting to be bored. They have already met the burned out detective, the last job before retirement, the interrogation room with the two way mirror. Learning how to write a crime screenplay is mostly learning which of those pieces are load bearing and which are decoration inherited from other people's movies. The genre punishes vagueness harder than almost any other, because its best examples are built on procedure that an audience can audit: how a tail gets made, how a lock gets bypassed, how a case actually files. Get the procedure right and the melodrama takes care of itself.

What Makes a Crime Screenplay Its Own Genre

Crime is a parent category, not a rulebook. It covers stories revolving around illegal activity and the institutions and people that respond to it. That is a wide net, which is why the word "crime" on a logline communicates almost nothing to a reader. Underneath it sit subgenres with genuinely incompatible conventions.

The caper story is separated from the straight crime story by humor, adventure, or unusual cleverness and audacity. Its protagonists are typically lovable rogues, and law enforcement in a caper is often characterized by ineptitude, because a competent police force would end the fun in twenty pages. The police procedural runs the opposite way: a documentary style approach that follows the investigative work of law enforcement, where competence is the pleasure. Film noir sits somewhere else again, defined by cynical attitudes and moral ambiguity, gloomy and shadowy settings, and hardboiled investigators moving through a corrupt underworld. Put caper style police incompetence into a procedural and you have not blended forms, you have broken one.

What holds the parent category together is consequence. In a crime story, an action produces a specific reaction from a specific institution, and the audience keeps a running ledger. A romance can get away with a coincidence in act three. A crime script cannot, because the audience has been trained by the first act to check your arithmetic.

The bar is also unusually high, because crime has produced some of the most studied screenplays in the business. When the Writers Guild of America polled its members for the "101 Greatest Screenplays" list, first published in 2006, the top of the chart was largely crime.

That is not meant to be intimidating. It is meant to tell you what a professional reader has in their head when they open your pages. They are not comparing your work to the average submission. They are comparing your handling of loyalty, institutions, and moral cost to the small number of scripts that defined how those things get dramatized.

How to Write a Crime Screenplay: Choosing Your Subgenre

Pick the subgenre before page one, and put the choice on the page in the first five minutes. Tone is a contract. A reader who spends ten pages calibrating whether your bank robbery is funny or terrifying is not reading your story, they are doing your job for you.

Four choices cover most of the territory. The heist film focuses on the planning, execution, and aftermath of a significant robbery, and its engine is procedural realism. The caper takes the same architecture and plays it for cleverness and audacity. The police procedural inverts the point of view and asks the audience to enjoy method rather than transgression. Film noir subordinates plot mechanics to atmosphere and moral compromise, which is why noir plots can be famously hard to summarize and still work perfectly on screen.

Hybrids are common and they are where a lot of the best crime thriller screenplay examples live, but hybrids only work when one form is clearly the spine. Fargo (1996), which won Joel and Ethan Coen the Academy Award for Best Original Screenplay at the 69th ceremony out of seven nominations, holds a comic register and a genuinely brutal consequence line at once, and never lets the comedy soften the body count.

Subgenre choice also decides what your research burden is. A caper can hand wave the forensics. A procedural cannot. If you are not willing to learn how a warrant is actually obtained in your jurisdiction, do not write the procedural, because the first reader who has watched a real one will close the script at the moment you fake it.

One more consideration: subgenres age at different rates, and the ones that feel exhausted are usually the ones being repeated rather than reworked. Emerald Fennell won Best Original Screenplay at the 93rd Academy Awards in 2021 for Promising Young Woman, the first woman to win that category in thirteen years, with a script that takes revenge thriller conventions and rebuilds them around a sexual assault driven plot. The conventions were not the problem. The unexamined use of them was.

Structure: How to Write a Crime Screenplay Scene by Scene

Crime screenplay structure is information management. In most genres, structure asks what happens next. In crime, structure asks who knows it, when they learn it, and what the gap between the audience's knowledge and the characters' knowledge is doing to the tension in the room.

That gap has two settings. Withhold from the audience and you get mystery: the reader is investigating alongside the detective. Give the audience more than the characters have and you get suspense: the reader watches a detective walk into a room and knows what is waiting. Both work. What does not work is drifting between them by accident, which is the most common structural failure in unproduced crime scripts. Decide, scene by scene, which one you are running.

Blake Snyder's "Save the Cat!" beat sheet, from his 2005 book, maps a script across fifteen beats from Opening Image to Final Image and is used and argued about across professional screenwriting. Treat it as a diagnostic rather than a mold. Its real value on a crime script is catching a dead midpoint, because crime midpoints are where the case usually needs to invert: the theory the detective has been building collapses, or the plan the crew has been rehearsing becomes impossible.

Nonlinear ordering is legitimate and it has to earn itself. Quentin Tarantino and Roger Avary won Best Original Screenplay at the 67th Academy Awards for Pulp Fiction (1994), which lost Best Picture to Forrest Gump. Christopher McQuarrie won the same award at the 68th Academy Awards in 1996 for The Usual Suspects (1995), a film that also took Best Supporting Actor for Kevin Spacey.

Scene by scene, apply one test. Every scene must move either information or risk, and preferably both. A scene where two cops discuss a theory that the audience already holds and nobody's exposure changes is a scene you can cut without a stitch showing. Cutting is a craft in its own right: the Coens won Best Adapted Screenplay at the 80th Academy Awards in 2008 for No Country for Old Men, part of a four award sweep that included Best Picture and Best Director, and they have described that adaptation as compression, cutting rather than inventing, with much of the dialogue taken verbatim from Cormac McCarthy's novel.

If your twist can be removed without the preceding ninety pages changing, it is not structure. It is a surprise, and surprises are cheap. The test is whether a second read is a different experience, because the audience now watches every scene knowing what was underneath it.

Building a Morally Compromised Protagonist

Crime protagonists work when their competence and their corruption come from the same place. That is the whole design. A detective's obsessive pattern recognition makes them brilliant at the case and impossible to live with. A crew leader's discipline keeps everyone alive and makes him incapable of walking away with the money.

The most studied crime screenplays tend to work this way. Their protagonists are rarely good people corrupted by circumstance. More often they are capable people whose particular talents find their fullest use in the wrong arena, or whose skill at uncovering the small crime makes them an instrument of the large one. Those scripts do not ask you to approve of the protagonist. They make you understand the logic.

The practical consequence for your pages is that you should stop writing scenes where a character debates the morality of what they are about to do. Those scenes read as an author reassuring the reader. Write the decision instead, made quickly, on grounds the character finds obvious and the audience finds appalling. Moral weight belongs in consequence, not in dialogue about consequence.

Backstory is the other trap. Crime scripts arrive stuffed with dead partners, dead wives, and dead children, all deployed to buy sympathy the writing has not earned. If the past event does not change how the character behaves in a specific scene, it is not backstory, it is an excuse note. Cut it and see whether anything falls over.

Antagonists get the same treatment. The strongest crime scripts give the opposing side a working method, not a personality disorder. If the detective's counterpart is smart, patient, and operating on a coherent set of rules, every scene between them has real stakes. If he is simply unstable, the only question left is when he will be caught.

Heist Mechanics: Plans, Crews, and the Double-Cross

Anyone asking how to write a heist movie script is really asking about three sequences: the plan, the execution, and the aftermath. The subgenre is built on procedural realism, bypassing alarms, tunneling, timing distractions, interwoven with interpersonal tension of the greed driven betrayal variety. Those two strands are not separate tracks. The mechanical problem and the human problem should be the same problem seen twice.

The plan sequence has one job: teach the audience the rules precisely enough that they can spot a deviation without being told. That means concrete geography, concrete timing, concrete failure conditions. Vague planning produces an execution sequence where nothing can go visibly wrong, because the audience has no baseline. Give the reader the floor plan and you can generate three minutes of unbearable tension out of a guard walking the wrong corridor.

Execution should deviate early and keep deviating. A heist that runs exactly as planned is a demonstration, not a scene. The useful discipline is to make every deviation a consequence of a character trait you established during the plan. The one who improvises, the one who freezes, the one who cannot leave money on the floor: each failure should be legible as that person, not as bad luck.

Crew design follows from that. Build each member as a capability paired with a liability, and make the liability the reason they were hired. The double cross then has to be planted long before it fires. A betrayal that arrives without earlier motivation reads as a plot requirement, and readers feel the machinery.

The aftermath is where most spec heist scripts collapse. Writers spend eighty pages on the job and ten on what it cost, which inverts the interest curve. The money is the least interesting thing in the third act. What the crew becomes to each other is the story.

Grounding helps enormously here, and the pedigree for it is real. Michael Mann's Heat came from a case: Chicago police officer Chuck Adamson pursued a robbery crew led by a man named McCauley in 1964, and after retiring he told Mann the story and gave permission for a screenplay based on it. Mann wrote the first draft in 1979. The film's coffee shop confrontation, the scene everyone remembers, is drawn from a real encounter between Adamson and McCauley, with dialogue reportedly recreated close to word for word from Adamson's own account. That is what research buys you: a scene nobody could have invented.

Noir Voice, Tone, and the Detective Point of View

Writing film noir on the page is a different technical problem than writing a heist, because the genre's identity lives in texture rather than mechanism. Noir thrillers are defined by cynical attitudes and moral ambiguity, gloomy and shadowy visual settings, and hardboiled investigators navigating corrupt underworlds. Every one of those elements has to survive translation into a document that a reader consumes in ninety minutes without a cinematographer attached.

The answer is not to direct the camera in your action lines. It is to write description in the voice of the world. Noir prose on the page can be selective and cold, noticing what a cynical observer would notice: what a room costs, who paid for it, what the light is failing to hide. Choose four details instead of twelve. The reader's imagination will do the shadows.

Point of view discipline is the other half. Classical noir binds the audience to the investigator, which means the audience learns nothing the detective does not learn and, critically, that the detective's misreadings become the audience's misreadings. If you break that binding to show the audience a scene the detective never witnesses, you have chosen suspense over mystery, and you should know you chose it.

Voiceover deserves one warning. Narration is a noir convention and it is also the fastest way to write a script that explains itself. If the narration is telling us what a character feels or what a scene means, it is doing repair work on scenes that are not carrying their own weight. If it is doing something the images cannot, ironic distance, compressed time, an admission the character would never make aloud, keep it.

Tone in noir is also a promise about the ending. The form's cynicism is structural: institutions do not get fixed, and the detective's win is usually partial and expensive. A noir that resolves cleanly reads as a different genre wearing the costume.

Format, Length, and the Page-to-Screen Math

Crime script format is standard screenplay format. There is no genre variant, and any script that invents one signals amateur before the first slug line. What crime does demand is fluency with a few specific tools: intercutting parallel action during a heist or a raid, montage for surveillance and planning, phone conversations, and time stamps handled cleanly rather than with a paragraph of explanation.

Length has a real ceiling and a citable one. The Academy Nicholl Fellowships require an original feature screenplay of 80 to 125 pages, in English, in standard screenplay format. That range is a fair proxy for what the wider market will read without prejudice. A 140 page crime spec is not ambitious, it is unread.

The page per minute rule of thumb is more useful than its critics allow and less precise than its believers think. A properly formatted 100 page screenplay tends to run roughly 95 to 110 minutes on screen. Empirical analysis of produced scripts puts the real average closer to 1.1 pages per minute, about nine percent more pages than runtime minutes, with substantial variation by genre and era, and fewer than one in five scripts landing on an exact one to one ratio.

Crime scripts skew because of what they contain. Dense procedural dialogue plays faster than its page count suggests. A tightly written heist sequence, all short action blocks and white space, plays far slower than its pages. If your script is 118 pages and half of it is action, expect a long film, and expect a reader who has developed an instinct for that to notice.

Getting a Crime Script Read, Registered, and Sold

Protect the work first, cheaply, then stop worrying about it. The U.S. Copyright Office electronic Standard Application fee is currently $65, with a discounted $45 electronic fee that applies when there is one author who is also the sole claimant and the work is a single work not made for hire. That is the common case for a solo spec screenwriter registering an original script. The Office has proposed raising the Standard Application to $85; the final fee schedule went to Congress on July 14, 2026, triggering a 120 day statutory review, so the higher fees are expected around mid November 2026 unless Congress objects. Today, the $65 and $45 figures still apply.

The Writers Guild of America West's script registry is the other option, at $20 for non members and $10 for members in good standing, valid for five years and renewable at the then current rate. It is a dated deposit, useful as evidence of when a draft existed, and it is not a replacement for federal copyright registration. Do the copyright registration.

For exposure, competitions remain the most legible route for a writer without representation. The Academy Nicholl Fellowships award up to five $35,000 fellowships each year to non professional screenwriters, and the program has been awarding them since 1986. Eligibility caps an applicant's lifetime motion picture and television writing earnings at $25,000 USD, and the submitted work must be an original feature screenplay. The PAGE International Screenwriting Awards has no standalone crime category, so crime thrillers, mysteries, and psychological thrillers are submitted under Thriller/Horror, with the option to enter the same script in additional genre categories for a further fee.

One last thing about what actually gets a crime script bought, which is neither structure nor formatting. It is authority. The French Connection (1971) won Ernest Tidyman the Academy Award for Best Adapted Screenplay and took five of its eight nominations including Best Picture, adapted from Robin Moore's 1969 nonfiction book about real NYPD narcotics detectives Eddie Egan and Sonny Grosso, fictionalized on screen as Popeye Doyle and Cloudy Russo. Egan himself appears in the film, playing his own character's boss. That is not a coincidence of casting. It is a signal of how close the material stayed to the people who lived it.

You do not need a retired detective to hand you a case. You do need enough specific, verifiable knowledge of your world that a reader stops auditing you and starts trusting you. In crime, that trust is the whole sale.

If WGA minimums are relevant to where your career is headed, they exist and are publicly posted at wga.org/contracts. Everything else on this list you can do this month, for under a hundred dollars, with a finished script and a clear head about which subgenre you actually wrote.