Flesch Reading Ease is a formula, developed by Rudolf Flesch in the 1940s, that scores a piece of text from 0 to 100 based on two things: average sentence length and average syllables per word. Shorter sentences and shorter words push the score up. Longer sentences and longer words push it down. That’s the entire mechanism. It has no idea whether the writing is any good.

That last part matters more for novelists than the score itself.

What the Score Measures

Nothing about voice, tension, character, or whether a sentence is doing its job. It measures two mechanical properties: how long your sentences are and how many syllables your words have. A technically simple sentence with no emotional weight at all can score identically to a technically simple sentence that’s devastating. The formula can’t tell the difference, because it isn’t looking at meaning.

Rough reference points, for context: scores in the 90s read as very easy, similar to a children’s book. Scores in the 60 to 70 range sit close to standard, comfortable adult prose. Scores below 30 indicate long, syllable-heavy sentences that read as dense, closer to academic or legal writing than commercial fiction.

Clear sentences of varied length naturally produce a score somewhere in the middle of that range, as a mechanical side effect of the formula itself, not because anyone is writing toward a number.

Use It as a Diagnostic

The mistake would be treating the score as a goal: rewriting sentences to push a number up or down. That produces flat, mechanical prose fast, because you’d be optimising for syllable count instead of for the sentence working.

What the score is useful for is spotting patterns you didn’t intend.

A chapter that scores unusually low compared to the rest of your manuscript might mean an exposition-heavy section has drifted into long, clause-stacked sentences without you noticing while you were deep in explaining something. Worth a look, not necessarily a rewrite, some information needs longer sentences to convey. But worth checking whether that density was a choice or an accident.

A chapter that scores unusually high might mean a section has gone choppy, short sentence after short sentence, in a way that wasn’t a deliberate pacing choice. Short sentences are a legitimate tool for tension and speed. A whole chapter of them by accident reads as thin rather than tense.

Consistency across a manuscript is more useful to check than any single number. A book that swings wildly between very high and very low scores chapter to chapter might have a voice that’s drifting, unless that variation is intentional, different narrators or timelines written with deliberately different registers.

Genre Changes What’s Normal

A thriller and a literary novel with unreliable, meandering narration will not, and should not, land on the same score. Commercial genre fiction tends toward shorter sentences and a higher score as a matter of readability and pace. Literary fiction has more latitude for long, complex sentences that a formula like this will always mark down, even when they’re doing exactly what they’re meant to do.

There’s no correct target score for a novel. There’s a correct score for what the book is trying to be, and the only way to know that is to have a sense of what similar books in your genre and space read like, not what the formula says they should.

Where It Falls Apart

The formula can’t detect: a sentence that’s grammatically simple but confusing because of unclear pronoun references. A sentence that’s long but reads easily because the clauses are well ordered. Dialect or invented vocabulary that inflates syllable counts without adding real difficulty. Deliberately fragmented prose used for effect, which will score as either very easy or oddly erratic depending on how the fragments land.

It’s one mechanical signal among many, worth a glance rather than something to chase.

Writing Desk tip: The AI Editor’s line edit lens already flags sentences that read as unusually dense or hard to follow as part of its chapter feedback. That’s the more useful signal than a single readability number, since it points at the actual sentence worth a second look rather than telling you a chapter’s average score changed.

Choosing writing software? See the best editing and proofreading software.