Skip to content
TextLab

Free Flesch Reading Ease calculator

Flesch Reading Ease with the working shown

TextLab's Flesch Reading Ease calculator puts a number between 0 and 100 on your writing using the formula Rudolf Flesch published in 1948, where 60 to 70 is plain English and anything under 30 is specialist prose. Free, no signup, and the score moves with every keystroke. Rather than handing you a bare number it prints the equation with your own word, sentence and syllable counts substituted in, separates the penalty for long sentences from the penalty for long words, and works out what each would have to become to reach 60.

  • 100% free
  • No signup
  • Equation shown in full
  • All seven bands
  • Not clamped at 100
out of 100higher is easier — 100 is a children's reader, 0 is a mortgage deed
0 · very difficult50100 · very easy

Your text through the equation

206.835 − 1.015 × (words ÷ sentences) − 84.6 × (syllables ÷ words)

206.835 − 1.015 × (0 ÷ 0) − 84.6 × (0 ÷ 0)

206.835 − 1.015 × 0.000 − 84.6 × 0.000

206.835 − 0.000.00 = 0.00

The seven bands

ScoreFlesch's labelSchool levelWhat lives there
90–100Very easy5th gradeA ten-year-old reads it without slowing down
80–90Easy6th gradeConversational — where dialogue and light fiction sit
70–80Fairly easy7th gradePopular magazines and most brand copy
60–70Standard8th to 9th gradePlain English; the band most style guides ask you to hit
50–60Fairly difficult10th to 12th gradeBroadsheet journalism and serious non-fiction
30–50DifficultCollegeAcademic articles and technical documentation
below 30Very difficultCollege graduateContracts, statutes and law reviews

Where your score sits

  • 65
    Reader's DigestFlesch's own benchmark for popular writing
  • 52
    Time magazineNews written for a general adult audience
  • 45
    Florida insurance policiesA statutory floor, not a style choice
  • 30
    Harvard Law ReviewWhere specialist prose sits

How to calculate Flesch Reading Ease

The arithmetic stays on screen throughout, so nothing is taken on trust.

  1. Paste a passage, not a sentence

    Flesch built and tested this formula on samples of 100 words, and below roughly that length the score swings absurdly — a single ten-word sentence of short words can score above 100. Paste a few paragraphs, or take three 100-word samples from the start, middle and end of a long document and score them separately, which is how the formula was meant to be used.

  2. Read the equation with your numbers in it

    The panel under the box shows the formula four times: as Flesch wrote it, with your word, sentence and syllable counts substituted, with those reduced to the two averages, and finally as the subtraction that produces your score. Seeing it laid out that way tells you immediately whether sentence length or word length is costing you more, because the two penalties are printed side by side.

  3. Pull the cheaper of the two levers

    There are only two things in the equation you can change. The page works out what your average sentence would have to shrink to for the text to reach 60 — the top of the plain-English band — with your vocabulary untouched, and what your average word would have to shrink to with your sentences untouched. Doing a little of each is nearly always easier than either alone.

Technical specifications

The formula206.835 − 1.015 × (words ÷ sentences) − 84.6 × (syllables ÷ words), exactly as Rudolf Flesch published it in 1948
ScaleNominally 0 to 100 and higher is easier, but unbounded in both directions: the English maximum is 121.22 and long technical prose can score below zero
BandsFlesch's own seven, from Very easy at 90 and above down to Very difficult below 30, each with the school level he attached to it
Working shownFour lines: the formula, your counts substituted, the two averages, and the final subtraction with each penalty separated
Sample sizeDesigned for 100-word samples; below about 100 words the score is unstable and above 100 it is stable enough to score a whole document at once
Syllable accuracyAbout 95% agreement with the dictionary on a held-out sample of 198 everyday words, using vowel-group estimation with corrections
What it cannot seeWord order, ambiguity, familiarity, jargon, missing context and whether the argument makes sense — it measures two lengths and nothing else
PriceFree, no signup, no sample limit, and the arithmetic is shown rather than hidden

Frequently asked questions

What is a good Flesch Reading Ease score?

Between 60 and 70 for writing aimed at a general adult audience — Flesch labelled that band Standard, and it corresponds to 8th to 9th grade schooling. Marketing copy and popular journalism often sit in the 70s; a broadsheet feature is comfortable in the 50s; anything below 30 is specialist prose that a general reader will have to work at. Aiming for 90 is not a virtue: writing that scores that high is usually written for children, and adult readers find it patronising rather than clear.

Can the score go above 100 or below zero?

Yes, both — the scale is a name rather than a range. The theoretical maximum for English is 121.22, reached by a one-word sentence made of a single-syllable word, and short simple text routinely scores above 100 in practice. There is no floor at all: a long sentence packed with polysyllables can drive the number well below zero, and dense legal or academic prose sometimes does. Some implementations, including the one in Microsoft Word, clamp the result at 100 and at 0, which is why an unusually easy passage can look identical to another when it is not.

Why does Microsoft Word give me a different Flesch score?

Because the two disagree about counting, not about the formula. Word applies its own syllable estimation and its own rules on what constitutes a sentence, and it clamps the output at 100. The most common source of divergence is structural: Word includes headings, list items and table cells in the count, and a bulleted list with no full stops is read there as one very long sentence. Score the running prose on its own in both places and the two land within a point or two of each other.

How exactly are syllables counted here?

By estimation, not by dictionary lookup — vowel groups counted, then corrected for silent final e, sounded -ed and -es, consonant plus -le, the -tion family and a short list of irregular words. On a held-out sample of 198 everyday words that agrees with the dictionary about 95 per cent of the time. The misses are compounds like homework, whose internal silent e cannot be seen from the spelling, and borrowings like recipe that sound a final e. No estimator does better than the low nineties without a full pronunciation dictionary, and the residual error moves a whole document's score by a fraction of a point.

Is Flesch Reading Ease the same as Flesch-Kincaid Grade Level?

No. They share an author's name, two inputs and nothing else. Reading Ease produces a number from roughly 0 to 100 where higher means easier; Flesch-Kincaid produces a US school grade where higher means harder, so the two move in opposite directions. They also use different coefficients, which is why you cannot convert one to the other with a formula, only approximate it. Flesch published Reading Ease in 1948; the grade-level version came out of a US Navy project with J. Peter Kincaid in 1975 to specify the reading level of technical manuals for enlisted personnel.

Does it work on text that is not in English?

No, and the number it produces for other languages is worse than useless because it looks plausible. The constants 206.835, 1.015 and 84.6 were fitted to English reading-test data, and the relationship between syllable count and difficulty is different in every language — German compounds and Finnish inflection wreck it, Spanish words average more syllables without being harder. Adapted versions with refitted constants exist for several languages, including Fernández Huerta's for Spanish and Amstad's for German, and those are what you want rather than this one.

Does any law actually require a Flesch score?

Several US states do, for consumer contracts. Florida requires most insurance policies to score at least 45 on Flesch Reading Ease, and comparable plain-language statutes elsewhere set their own floors for policies and consumer agreements. That is unusual: almost everywhere else the score is a guideline rather than a requirement. It does mean the formula occasionally has to be computed exactly rather than approximately, which is why the working is printed above instead of just the answer.

The Austrian lawyer who taught America to write short sentences

Rudolf Flesch arrived in the United States in 1938 as a refugee with a law doctorate from Vienna and no English worth speaking of. He learned the language, took a doctorate in educational psychology at Columbia, and spent the rest of his career arguing that most professional writing is unnecessarily hard on purpose. The Art of Plain Talk appeared in 1946; the Reading Ease formula followed in 1948 in the Journal of Applied Psychology, fitted against reading-comprehension test data of the period. He is better known to the general public for Why Johnny Can't Read, the 1955 book that started the American phonics wars, but it is the 1948 paper that ended up embedded in software. The US Navy adapted it with J. Peter Kincaid in 1975 so that training manuals could be specified by reading grade, several US states wrote a minimum score into their insurance statutes, and word processors have shipped it since the 1990s — which is how a formula fitted to mid-century schoolchildren came to sit in the status bar of every document written since.

The formula itself is disarmingly simple, and the two coefficients tell you what Flesch believed. Sentence length is multiplied by 1.015 and syllables per word by 84.6, so on the scale that emerges a word gaining a syllable costs far more than a sentence gaining a word. In practice sentence length usually dominates anyway, because average sentence lengths vary between about 10 and 40 while syllables per word barely move outside 1.3 to 1.9 — the panel above prints both penalties so you can see which is true of your text rather than guessing. The constant 206.835 exists to put the result on a 0-to-100 scale for typical prose; it does not stop the score exceeding 100 or falling below 0, and a calculator that clamps it is hiding information.

What the formula cannot do is worth knowing before you quote it. It never reads a word. Shuffle the words of a sentence into nonsense and the score does not move, because the syllables and the full stop are all it can see. It rates chthonic easier than relationship, since one has two syllables and the other four. Names, acronyms and numbers distort it, and a list without full stops reads as one enormous sentence. None of that makes it useless — across a whole document the correlation with real difficulty holds up — but it does mean the score is evidence, not a verdict. For the same text seen through five other formulas and a median grade level, the readability score panel is the wider view; the syllable estimates driving both come from the same code as the syllable counter, and if your sentence lengths look like the problem the sentence counter will show you which one ran away with itself.

The passage you score is never transmitted

Nothing here requires a server: the syllable estimation, the counting and the three multiplications all run in the page in front of you. That is what makes it reasonable to paste a draft insurance policy, an unpublished patient leaflet or somebody else's coursework into the box — the text is not uploaded, not kept and not readable by anyone but you. Refresh and both the passage and its score are gone for good.