How to Write Action Lines in a Screenplay

The rules: present tense, active voice, third person

Action is written in the present tense, no exceptions. Not past for flashbacks, not future for what a character plans to do, not conditional:

GOOD:  MARLOWE crosses the room and opens the window.
BAD:   Marlowe crossed the room and opened the window.
BAD:   Marlowe will cross the room and open the window.

Flashbacks are the place writers break this rule, usually because past tense feels more natural for something that already happened. It is still present tense in the dominant convention: the camera is running in the past, but it is running now. The tense is a property of the screenplay, not of the scene's chronology.

Use active voice, with a person or thing as the subject of the verb doing the work:

GOOD:  Blood seeps through her sleeve.
BAD:   Her sleeve is stained with blood.
GOOD:  The window shatters inward.
BAD:   The window is broken by something from outside.

Third person only. No first person narration, no direct address to the reader, no I and no you — a script that talks to its reader is a treatment wearing a script's margins.

Fourth, and the one that decides everything else: action describes only what can be photographed or recorded. Not what a character thinks, not what they remember, not what the audience is meant to understand, not how the air smells unless the smell is visible as a physical fact:

NOT FILMABLE:  The room smells of bleach and something was cleaned up in a hurry.
FILMABLE:      Bleach. Buckets stacked against the wall. A wet mop abandoned in the middle
               of the floor.

The second version also happens to be shorter than the first, which is the usual result of converting abstraction into images. Verbs are your cheapest tool: limps, flinches, pockets the key carry characterisation that an adjective cannot.

Write what the camera can see and the ear can hear

The reader is not imagining a novel. They are running a film in their head at something close to reading speed, and their sense of the scene's pace comes from the shape of the text as much as from its content. Two consequences follow.

Write behaviour, not interiority. If the scene needs the audience to know that a character is afraid, write what fear looks like from outside, and write the beat where it becomes visible:

GOOD:  Her hands are steady. Her eyes are not.
BAD:   She is terrified of what happens next.

Write the sound, not the feeling. Sound is half of what the page can carry — a television through a wall, a fridge cycling off, a lock turning — and it is the cheapest way to imply an offscreen world. But write it as a fact the microphone records, not as atmosphere: A television murmurs through the wall is filmable. The silence is oppressive is not.

One more discipline that separates working action from amateur action: describe the event, not the shot that captures it. The director and the editor decide where the camera goes. You decide what happens in front of it, and the reader will see it from the best angle anyway.

Paragraph length, white space, and pace

Action is broken into short paragraphs, one to four lines each, separated by a blank line. The widely repeated guideline is three lines maximum, and while no rule is enforced by anything, the reason behind it is real: reading speed is the reader's only signal for screen pace.

A page of dense block text reads slowly, so the scene feels slow even if what it describes is a fight. A page of short paragraphs reads quickly, so the scene feels fast. You are not decorating the page — you are calibrating it.

GOOD (three beats, three paragraphs):

He shoulders the door. It holds.

He shoulders it again. The frame splinters.

Inside: nothing but a chair, and the chair is facing the door.

Practical rules that follow:

• One beat, one paragraph. When the subject or the action changes, start a new paragraph.
• A single-line paragraph is a hard beat. Use it for a reveal or a hard cut. Use it on every line and all your beats weigh the same, which means none of them land.
• Never run action past five or six lines. If a paragraph is that long, it is usually two beats stuck together with a comma, and splitting it will improve the scene.
• Blank lines are not free. Standard format gives you about 55 lines on a page, and every blank line spends one of them. Generous white space is good writing; padding a page with blank lines to reach a page count is not.

Dialogue gets the same treatment for the same reason. Four lines of speech from one character is a paragraph of screen time; ten lines needs a reason to exist, and usually that reason is that the speech is the scene.

Capitalising props and sounds

Caps are emphasis, and emphasis is a limited resource. Two things earn it, and only two.

Sounds — capitalised when we hear the sound before we see its source, when it is objectively loud, or when it matters to the scene:

A GUNSHOT. He drops to the floor before the window finishes breaking.

Somewhere below, a DOOR SLAMS.

Props and objects — capitalised the first time they appear when they carry meaning, because the caps are a note to the props department and to the reader:

On the table: a REVOLVER, a set of keys, and a photograph face down.

That is the whole convention. The rule of thumb working writers apply is a few caps per page and no more; a new prop gets caps once, at its introduction, and never again.

Over-capitalising is the most common way to kill the effect. If every noun is shouting, nothing is emphasised and the page reads as loud and amateurish:

TOO MUCH:  He walks to the DOOR and opens it. The CAR in the DRIVEWAY is a RED SEDAN.
RIGHT:     He opens the door. Through it: a red sedan in the driveway, engine running.

Things that do not get capitals: ordinary nouns (door, car, phone), every verb in a violent scene, character names after their introduction, and words you simply want the reader to feel. Caps are not italics. If a word only needs emphasis, the sentence needs rewriting.

Why 'we see' marks a spec script

We see a car pull into the driveway is the second sentence of a shooting script or the first sentence of a script by someone who has not read many scripts. In a spec script the dominant practice is to delete we see, we hear, we watch, and we notice entirely:

BEFORE:  We see a car pull into the driveway. We watch as she gets out and we
         realise she has not been here in years.
AFTER:   A car pulls into the driveway. She gets out and stands by it a beat
         longer than she needs to.

There are two reasons. The mechanical one is that the camera sees whatever the action line describes — the phrase adds nothing. The practical one is that we is a filter between the reader and the story, and the reader is looking for an excuse to stop reading. A page of we reads as a writer explaining the movie instead of showing it.

The same logic removes camera directions from a spec: ANGLE ON, CLOSE ON, CAMERA PANS, WIDE SHOT, POV. These are production language and belong in a shooting script where the director and the storyboard artist are already part of the conversation. In a spec they are evidence that the writer is trying to direct.

This is one of the genuinely contested points in screenwriting craft. Some produced writers use we as a deliberate voice device, and a script with a strong voice can survive it. The dominant position among readers, contests, and script consultants is still to cut it — so if you keep it, keep it as a decision you can defend rather than an accident of habit.

Two related habits go with it. Do not write what the audience realises — write the moment they realise it. And do not write what a character is thinking: She knows he is lying is a novel sentence. She looks at him a beat too long, then nods is a script sentence.

Action for montage and for silence

Montage. There are two accepted formats. The labelled version brackets the sequence:

BEGIN MONTAGE

Boxes. Tape. A mattress dragged down a stairwell.

Maya eats standing up, over the sink.

Each night, one fewer box in the apartment.

END MONTAGE

The titled version uses a heading and a list of beats, which is the more common modern shorthand:

MONTAGE - MAYA MOVES OUT

- Boxes stacked three deep.
- A mattress dragged down a stairwell.
- Maya eats standing up, over the sink.
- Each night, one fewer box.

Either way, one line is one shot. That is why montage pages play far faster than the page-per-minute average: each line is a second or two on screen. Keep the beats present tense, keep them to a single image, and let the last one land the idea — a montage is an argument, not a highlight reel.

Silence. Write the silence by writing what is physically happening, and let the missing dialogue be the silence. A page or two without dialogue is a scene carried by behaviour, so the action has to be about the characters rather than the room:

She reaches for the salt. He moves it out of reach.

Neither of them eats.

Two useful devices when a scene has to be quiet for a long stretch. First, a single capitalised sound in an otherwise silent passage cuts through and marks the beat — a CLOCK TICKING, a fridge hum, a floorboard. Second, a short line of three or four words alone in a paragraph reads as a held beat, which is what a silence on screen feels like. Do not annotate the silence with notes like (no dialogue); the empty space on the page already says it, and telling the reader how to feel about the quiet is the same mistake as we see.

Five action rewrites, before and after

1. Abstraction into images. Same information, filmable, and shorter:

BEFORE:  The apartment has clearly been abandoned for years, with dust everywhere and
the sense that nobody has lived here in a very long time.

AFTER:   Dust on everything. A calendar on the wall, open to a page three years old.

2. Cutting 'we'. The camera already sees it:

BEFORE:  We see her hesitate at the door, and we understand that she has not been
here since the funeral.

AFTER:   She stops at the door. Wipes her palms on her coat. Goes in.

3. Passive into active. Find the thing doing the verb:

BEFORE:  The window is shattered from the outside, and glass is scattered across the
floor where it has fallen.

AFTER:   The window blows inward. Glass skitters across the floorboards.

4. A wall of text into beats. Three sentences, three paragraphs, three visible actions:

BEFORE:  He walks into the kitchen, opens the fridge, finds it empty, closes it again,
then sits down at the table and puts his head in his hands, exhausted by the
realisation that there is nothing left in the house to eat.

AFTER:   He opens the fridge.

Empty shelves. A single jar, lid off.

He sits down and puts his head in his hands.

5. Over-capitalising and dead weight. One caps sound survives because it is the loud thing; nothing else earns caps:

BEFORE:  She WALKS into the BEDROOM and SLAMS the DOOR. A CAR ALARM goes off outside
and she looks out of the WINDOW at the STREET below.

AFTER:   She slams the bedroom door. Through the glass: a car alarm going, off and on,
off and on.

Every after version is doing three things at once: present tense, an actor's behaviour, and a reader's pace. That is the whole job of an action line.

Frequently asked questions

Continue reading

More in screenwriting education

• What Is a Logline? Examples, Formula & Free Generator
• How to Write a Logline That Sells Your Script
• Logline Examples: 20 Strong and Weak Loglines Compared
• What Is a Beat Sheet? Save the Cat and 3-Act Structure Explained
• What Is a Treatment? Screenplay Treatment Format Explained
• What Is a Screenplay? A Complete Beginner's Guide
• How to Write a Screenplay: A Step-by-Step Guide
• How to Write a Short Film
• How Long Is a Screenplay? Page Counts and Runtime
• How Many Screenplay Pages Equal One Minute?
• How to Write Character Introductions in a Screenplay