Should You Write the Ending First?
Quick answer: Should you write the ending first? When knowing or drafting your final scene strengthens a story, which writers work backwards, and when discovery serves you better.
You can, and for certain stories you probably should: writing the ending first, or at least knowing it precisely, gives every earlier scene a target, makes foreshadowing deliberate instead of lucky, and is close to mandatory for twist stories and mysteries. John Irving famously writes his last sentence before anything else. But it is a tool, not a law: plenty of fine writers discover endings in the drafting, and an ending fixed too early can handcuff a story that wanted to become something better.
The question matters because endings are where stories are judged, and where drafts most often fail. Deciding when you will invent yours, first, midway or last, is one of the few process choices that changes the finished story's shape, not just the writing experience.
The longer answer: what ending-first actually buys
Causality is the core purchase. Stories read as inevitable when every scene quietly pays towards the conclusion, and that is far easier to engineer when the conclusion exists. Writing towards a known ending converts drafting from exploration into aiming: each scene either moves the story closer or gets cut, and the middle, fiction's swampiest terrain, acquires gravity. Writers who stall at sixty percent are almost always writers without a destination.
The second purchase is foreshadowing at cost price. Twists, reveals and resonant final images all depend on planted material, and planting is trivial when you know the harvest: the mystery writer seeds the clue in chapter two because chapter twelve already exists, at least on a card. Working backwards from effect to cause is standard practice in crime writing for exactly this reason, and it is why fair-play mysteries are nearly impossible to pants twice.
What ending-first costs is optionality. Drafts learn things, characters deepen, better endings sometimes surface en route, and a writer too loyal to the pre-written finale will march past superior material to keep an appointment. The mature version of the method holds the ending firmly and the route loosely, and renegotiates the destination on the rare occasions the draft presents a genuinely stronger one.
Writers who work backwards, and writers who refuse
John Irving is the standing example: he has said repeatedly that he begins with the final sentence, sometimes carrying it for years before drafting towards it, and that for him a novel is the story of everything that must happen to earn that sentence. Katherine Anne Porter claimed she always wrote her last lines first, insisting that without knowing the destination she would never start the journey. Joseph Heller often generated books from a first and last conception early, and mystery writers from Agatha Christie onward have routinely built whole plots backwards from the solution; Christie's notebooks show her testing culprits and then reverse-engineering the clues.
The refusals are equally distinguished. Stephen King treats stories as found objects and considers pre-set endings a form of cheating against the characters. George R. R. Martin's gardener metaphor implies endings that grow rather than get installed. Between the camps sits a large pragmatic majority who know the ending's direction, who wins, what it costs, what the final image feels like, without drafting its prose, keeping the target's gravity while leaving its wording free. For short fiction especially, that middle position is common professional practice: the closing image on an index card, the route discovered.
How to try it on your next story
Before drafting scene one, draft the final scene, or if that feels too binding, write three sentences: who is present, what has changed, and the final image. Pin it where you can see it. Then draft forward normally, testing each scene with one question: does this move the story towards the card? You will feel the difference immediately in the middle, where scenes stop auditioning and start working.
Use the full ending-first treatment when the story's effect depends on engineering: twists, mysteries, tragedies where the fall must feel inevitable, and any story with a hard word limit, because knowing the ending prevents the classic contest failure of a leisurely start and a crushed finale. Under a 2,000-word ceiling, the kind Writory's monthly story contest sets, backwards construction is close to a competitive advantage: judges scoring structure can tell which stories knew where they were going.
Grant yourself one renegotiation clause: if the draft produces an ending demonstrably stronger than the card, take it, then revise earlier scenes so the new ending is as earned as the old one would have been. The method's point is an earned ending, not a kept appointment.
Common mistakes
Mistake one: confusing knowing the ending with writing it verbatim too early. Pre-written final pages often arrive embalmed, since they were composed before the story's voice matured; the card with three sentences usually serves better than two polished pages. Mistake two: loyalty to a dead destination. If every scene since the midpoint has fought the planned finale, the draft is voting; count the votes.
Mistake three: ending-first as spoiler-fear. Some writers avoid deciding endings because a known ending feels less exciting to write. But the reader's suspense does not depend on the writer's ignorance; it depends on the page. Craft manufactures surprise reliably, luck does not. Mistake four: aiming at an ending without an emotional target, plotting the event, he confesses, without the feeling, relief or ruin? The card should name both. And mistake five: assuming the method suits every temperament. If a known destination reliably kills your drafts, you have learned something real about your process; keep the direction vague and the compass emotional, and let the last scene stay dark until you arrive.
What poets can take from this
- Write the ending as a three-sentence card: who is present, what changed, final image.
- Name the ending's emotion, not just its event.
- Test every middle scene against the card; scenes that do not move towards it audition for deletion.
- Go fully backwards for twists, mysteries and hard word limits.
- Allow one renegotiation if the draft finds a stronger ending, then re-earn it in revision.
- If fixed endings kill your drafts, keep a direction instead of a destination.
Try it yourself
Write the final 150 words of a story first: two estranged friends, one returned key, whatever else you like. Then write the 1,200 words that earn those final lines exactly as drafted, changing nothing in them.
Questions
Do mystery writers really plot backwards?
Very commonly, yes. The solution, who did it, how, why, is designed first, then clues, red herrings and alibis are laid backwards through the timeline. Agatha Christie's surviving notebooks show her selecting the culprit and reverse-building the puzzle around them.
What makes a story ending feel earned?
Causality and payment: the ending uses only materials the story visibly planted, and the protagonist's final state follows from their choices, not from coincidence or a new character. Surprising yet inevitable is the standard formula, and inevitability is manufactured earlier, not at the end.
Is it bad luck to tell people your ending before writing it?
Not luck, but there is a real effect: narrating a story aloud can discharge the urge to write it, and some writers guard endings for that reason. If telling drains you, keep the card private; if telling energises you, it costs nothing. Know which writer you are.
By the Writory editorial team, reviewed by working poets. Updated July 2026.
Enter a Writory poetry contest
Enter the August contest and have your own poem read and scored.