In October 1928, Dorothy Parker reviewed The House at Pooh Corner for The New Yorker. She got as far as the moment Pooh calls his song "more hummy," and stopped. Her verdict came in baby talk: "Tonstant Weader Fwowed up."
People still quote that line almost a century later. Most reviews vanish by the next morning.
The gap between the two has little to do with talent. Parker had a clear opinion and a moment in the book to pin it on. She also had a voice you could hear. Anyone can learn to do the same.
This guide covers how to write a book review from the first line to the last. You'll get a five-part structure, seven ways to open, and a method for negative reviews that stay fair. Each step comes with an example from a critic who did it well, or badly enough to teach something.
What a book review is for
A book review answers three questions for someone who hasn't read the book. What is it? Is it any good? Should I read it?
A summary only answers the first. A school book report proves you read the thing. A review goes further: it makes a judgment, then backs it up.
Kirkus Reviews, one of the best-known trade review outlets in the US, builds every review on that split. Its indie review service describes each one as summary plus opinion-based analysis. A standard review runs about 250 words.
That's the whole job, done in a small space. Everything below is about doing it well.
The five-part structure of a book review
Most good reviews move through the same five parts, from a 150-word app review to a 6,000-word magazine essay. The proportions change with length. The order rarely does.
1. The hook
Your first sentence decides whether anyone reads the second. Open on your opinion, a scene, a question, or a line from the book.
Avoid opening on the title and the author's name alone. That reads like a library catalogue entry. You'll find seven hooks that work further down.
2. The context
Next, tell readers what they're holding: title, author, genre, and where it sits in a series or a career. Add the length if it matters.
Then say what the book is trying to do. John Updike made this his first rule of reviewing, in his 1975 collection Picked-Up Pieces. Work out what the author wanted to do, he wrote. Then don't blame them for missing a target they never aimed at. The National Book Critics Circle keeps his full list.
In practice, a cozy mystery written to comfort shouldn't lose marks for lacking a thriller's menace.
3. A short summary, without spoilers
Give just enough plot for readers to grasp the premise. Who is the story about? What do they want, and what stands in the way?
For most novels, that means the setup and nothing past the first act. Updike's advice was to go easy on plot summary and never give away the ending. A handy ceiling: say no more than the jacket copy says.
Nonfiction works the same way. Name the argument and the kind of evidence the author uses, then move on.
4. Your assessment, with evidence
This part is the review. It's also the part most people skip, jumping from the plot straight to "five stars."
Pick two or three things the book does well or badly, and prove each one. Quote a sentence. Describe a scene. Point to the chapter where the pacing sagged.
Updike asked reviewers to quote at least one long passage, so readers can taste the prose for themselves. Compare these two lines about the same novel:
- "The writing is beautiful."
- "Fitzgerald can make one sentence carry a whole party: the music, the gin, and the loneliness of the man paying for it."
The first is an opinion. The second is an opinion with proof attached, and it tells readers what kind of beautiful to expect.
5. The verdict, and who it's for
End with a clear judgment. A star rating works, and so does a letter grade or one plain sentence.
Then name the reader. Stephen King reviewed The Hunger Games for Entertainment Weekly in 2008 and gave it a B. He also warned that fans of Battle Royale, The Running Man and The Long Walk would recognize the territory. He wrote the last two himself, under the name Richard Bachman.
That one comparison told a reader more than the grade did. "Read this if you loved The Secret History" does more work than "highly recommended."
How the five parts scale
| Part | 250-word review | 1,000-word review |
|---|---|---|
| Hook | 1 sentence | 1 short paragraph |
| Context | 1 or 2 sentences | 1 paragraph |
| Summary | 2 or 3 sentences | 2 paragraphs |
| Assessment | 3 or 4 sentences, 1 quote | 4 to 6 paragraphs, several quotes |
| Verdict | 1 or 2 sentences | 1 paragraph |
Watch where the extra space goes. As a review grows, the summary barely changes. The assessment takes almost all of it.
Seven ways to open a book review
A hook promises that the review has a point of view. Here are seven that work, most with a real example and each with a sample line you could adapt. The samples all use The Great Gatsby, so you can compare them side by side.
1. The flat verdict
Say what you think in the first sentence. Nothing grabs faster.
Dale Peck opened his 2002 New Republic review of Rick Moody's memoir with "Rick Moody is the worst writer of his generation." It became one of the most talked-about first lines in modern criticism. It also followed Peck for years, and he later titled a collection of his reviews Hatchet Jobs.
Use this hook only when you can defend the claim for the next thousand words.
Try it: "The Great Gatsby is the best short novel anyone has written about American money."
2. The borrowed voice
Parker's Pooh review is the model here. She wrote her verdict in the book's own register, so the joke and the criticism landed as one line. Here is the full sentence from her Constant Reader column of 20 October 1928:
"And it is that word 'hummy,' my darlings, that marks the first place in 'The House at Pooh Corner' at which Tonstant Weader Fwowed up."
Voice hooks carry risk. You need a book whose register is worth borrowing, and a point that survives once the joke is over.
Try it: "Old sport, you'll want to read this one twice."
3. The scene
Drop readers into a moment from the book, then pull back to your judgment. The scene lends its pull to your review.
Try it: "A man stands alone on his lawn at night, reaching toward a green light across the bay. By the end of chapter one, Fitzgerald has shown you his whole hand, and you keep reading anyway."
4. The quoted line
Open with a sentence from the book itself. When the prose is the reason to read it, let the prose make the case.
Try it: "'In my younger and more vulnerable years my father gave me some advice that I've been turning over in my mind ever since.' The rest of the novel shows how badly the narrator follows it."
5. The personal frame
Tell readers how you came to the book, but only when it changes how they should weigh your verdict.
George Orwell opened his 1946 essay Confessions of a Book Reviewer with the reviewer himself. He sits in a cold, untidy bedsit in a dressing gown. Waiting for him is a parcel of books he doesn't want to read. The essay is about reviewing, but the move carries over. Readers trust a critic who admits the conditions.
Try it: "I first read Gatsby at sixteen and thought it was a love story. At forty, it reads like a bank statement."
6. The argument about the label
Open by questioning what kind of book this is. Ursula K. Le Guin did this in her 2009 Guardian review of Margaret Atwood's The Year of the Flood.
Atwood refuses to call her novels science fiction. Le Guin spent her opening on that refusal, and on the false position it puts a reviewer in. Only then did she turn to the novel.
Try it: "Schools teach Gatsby as a love story. Read it as a novel about class, and every party scene changes."
7. The comparison
Set the book beside one your reader already knows. King's Battle Royale line did this in a single sentence.
Try it: "If you loved the outsider narrator of The Secret History, meet his great-grandfather."
What to judge: the questions behind a good assessment
"Four stars" with no reasons tells the next reader almost nothing. Breaking the book into parts gives you something specific to say.
For fiction, ask:
- Writing: does the prose do its job, or does it get in the way?
- Characters: do they want things, and would they hold a scene without the plot shoving them?
- Plot: does it move, and could you see the turns coming?
- Atmosphere: can you feel where you are?
- Impact: is it still with you a week later?
For nonfiction, swap three of those out:
- Ideas: is the central argument new, or a magazine article stretched to 300 pages?
- Logic: does each chapter build on the last?
- Clarity: could a smart outsider follow it?
You don't need to cover all of them. Pick the two or three where the book is strongest or weakest, and spend your words there.
Twain's rules for judging prose
For a sharper lens on the writing itself, borrow from Mark Twain. His 1895 essay "Fenimore Cooper's Literary Offences" is a review built as a rulebook. He lists 18 rules of fiction and shows Cooper breaking nearly all of them. Three still make good tests for any book:
- "Say what he is proposing to say, not merely come near it."
- "Use the right word, not its second cousin."
- "Eschew surplusage."
The essay is also a lesson in evidence. Twain claims that in two-thirds of a page of The Deerslayer, Cooper scored "114 offences against literary art out of a possible 115." It's a joke. It works because he has quoted enough Cooper by then to make it believable.
Poe's test for story collections
Reviewing short stories? Edgar Allan Poe set the classic test in his 1842 review of Nathaniel Hawthorne's Twice-Told Tales. A good tale, he argued, aims at a single effect, and every sentence should serve it.
Ask that of each story in a collection. What was it going for, and did every page pull that way?
How to write a negative review that's fair
Negative reviews are the hardest to write well and the easiest to write badly. Your target is the book, never the person who wrote it.
From 2012 to 2014, the review site The Omnivore handed out a Hatchet Job of the Year award. It went to the "angriest, funniest, most trenchant" review of the previous year. Winners included Adam Mars-Jones on Michael Cunningham's By Nightfall and Camilla Long on Rachel Cusk's memoir Aftermath. The prize was a year's supply of potted shrimp.
The strongest takedowns share a method. They quote the book and let its sentences fail in front of the reader. Anger without evidence reads as a grudge.
Updike's fifth rule helps here. If a book falls short, name one that got the same thing right. It can come from the author's other work or anyone else's. Pair a complaint with a comparison and it turns into criticism: "The twist lands with a thud. Compare how Gone Girl hides its hinge in plain sight."
Two critics, one book, opposite verdicts
In 2015, Hanya Yanagihara's A Little Life drew two of the most discussed reviews of the decade. In The Atlantic, novelist Garth Greenwell suggested it might be the great gay novel of its moment. In The New York Review of Books, Daniel Mendelsohn tore into its prose and its handling of suffering. The book's editor published a reply objecting to the review.
Both critics argued from the text. That's why readers still pass both reviews around, and why anyone can weigh them and decide for themselves. A fair negative review leaves room for that.
Updike had one more rule, about the reviewer. Don't take on a book you expect to dislike, or one a friendship commits you to like.
Spoilers: where the line sits
Updike's rule was blunt: don't give away the ending. Most reviewers now stretch that to any turn the jacket copy leaves out.
Even professional critics slip. In July 2007, Michiko Kakutani reviewed Harry Potter and the Deathly Hallows for The New York Times. The review ran before the book came out. She had bought a copy from a store that broke the embargo. J.K. Rowling publicly criticized the Times and a second paper that reviewed it early.
A few rules of thumb:
- Treat everything past the first act as off limits unless you hide it.
- Don't announce that a twist is coming. For some books, that alone gives it away.
- If you must discuss the ending, warn readers first and use a spoiler tag.
- In a series review, assume readers haven't finished the earlier books.
Most reading apps let you hide spoilers behind a tap. Seekquel's review editor lets you mark a single passage as a spoiler. You can also set how much the whole review gives away.
How long should a book review be?
As long as it takes to prove your verdict, and no longer. Here are typical lengths by format:
| Where it runs | Typical length | What fits |
|---|---|---|
| Reading app or retailer | 100 to 300 words | Hook, one-line summary, two reasons, verdict |
| Kirkus standard review | About 250 words | Tight summary, one clear judgment |
| Kirkus expanded review | About 500 words | Summary plus fuller analysis |
| Book blog | 600 to 1,200 words | All five parts, several quotes |
| Magazine essay | 2,000 to 6,000 words | The book, the author's career, a wider argument |
Peck's Moody review ran close to 6,000 words. Most readers will never need that much room.
A short review still needs a verdict and one reason. "Loved it!!!" helps nobody who's deciding.
Disclose free copies
Did a publisher or author send you the book? Say so near the top, in plain words: "The publisher sent me a free advance copy."
In the US, the FTC's endorsement guidance covers this. Reviewers should disclose when a company gave them something of value, like a free book, to write about. A friendship with the author deserves the same mention.
A worked example: a 250-word review
Here is a short review of The Great Gatsby that uses all five parts. Read it through once, then check the breakdown below it.
Schools teach The Great Gatsby as a love story. Read it again as an adult, and it turns out to be a book about who gets to belong.
F. Scott Fitzgerald's 1925 novel is short, under 200 pages in most editions. Its narrator is Nick Carraway, a young bond salesman from the Midwest who rents a cottage on Long Island. His neighbor, Jay Gatsby, throws enormous parties and seems to know nobody at them. Across the bay lives Nick's cousin Daisy, and Gatsby wants something from her that Nick slowly works out.
The prose is the reason to read it. Fitzgerald can make one sentence carry a whole party, from the music and the gin to the loneliness of the man paying for it. His guests come and go like moths. Nick watches from half a step away, and his distance becomes yours.
The book falls short on Daisy. She's more symbol than person, seen only through two men who want her. Set her beside Lily Bart in Edith Wharton's The House of Mirth, a woman with wants of her own, and the gap shows.
Four and a half stars. Read it if you liked the outsider narrator of The Secret History, or if you only met Gatsby in a classroom. Few novels this size reward a second reading so well.
Here's how the parts break down:
- Hook: the first two sentences argue about the label.
- Context: author, year, length, narrator and setting, in two lines.
- Summary: the setup only, with nothing past the early chapters.
- Assessment: one strength and one weakness, each backed by evidence or a comparison.
- Verdict: a rating, then two kinds of reader it suits.
Your checklist before you post
- Does the first sentence carry an opinion, a scene, a question or a quote?
- Could a stranger tell what the book is about from your review alone?
- Is everything past the first act either cut or hidden?
- Does each claim have a quote, a scene or a comparison behind it?
- Did you judge the book the author wrote, rather than the one you wanted?
- Does the last line say who should read it?
- If you got a free copy, did you say so?
Start the review the night you finish
The easiest review to write is the one you start while the book is fresh. Jot two lines the night you finish, while the scene that got you is still sharp. Shape them into the five parts the next day.
Seekquel keeps that draft next to the book in your library. Alongside your star rating, you can score a novel's writing, characters, plot, atmosphere and impact out of five. Nonfiction gets ideas, logic flow and clarity instead. Those scores