Flesch-Kincaid Readability Calculator: Formulas and Fixes

November 11, 2026 · 13 min read

Flesch-Kincaid Readability Calculator: Formulas, Scores and Fixes

A Flesch-Kincaid readability calculator counts your words, sentences and syllables, then turns them into two scores. Reading Ease runs from 0 to 100, and higher is easier. Grade Level is a US school grade, and higher is harder. This guide works both formulas by hand, compares five other readability formulas, and shows the edits that move the score.

What a Flesch-Kincaid Readability Calculator Measures

A readability score is a number that estimates how hard a passage is to read, based only on how long its sentences and words are. It doesn't read for meaning. It counts.

Every Flesch-Kincaid readability calculator needs three counts:

  • Words, split on spaces.
  • Sentences, usually split on full stops, question marks and exclamation marks.
  • Syllables, estimated from vowel groups, because no calculator carries a full pronunciation dictionary.

From those counts it gets two averages. Average sentence length (ASL) is words divided by sentences. Average syllables per word (ASW) is syllables divided by words. Both Flesch formulas use the same two inputs with different weights.

The two scores still move in opposite directions. Reading Ease subtracts the averages from a constant, so long sentences push it down. Grade Level adds them up, so long sentences push it up. A Reading Ease of 80 and a Grade Level of 4 describe the same easy text.

You can paste a draft into our readability checker to get both numbers without doing the arithmetic. It runs in your browser and also reports Gunning Fog, SMOG, Coleman-Liau and ARI.

How Is the Flesch Reading Ease Score Calculated?

Rudolf Flesch published the Reading Ease formula in 1948. It's the older of the two Flesch scores and the one with the 0 to 100 scale. When someone quotes a Flesch readability score without saying which, they usually mean this one.

Reading Ease = 206.835 - (1.015 x ASL) - (84.6 x ASW)

The syllable term does most of the work. Its weight is 84.6, against 1.015 for sentence length. Adding half a syllable per word costs about 42 points, while adding one word per sentence costs about 1.

The score bands below follow the table usually printed alongside Flesch's formula, as summarised on Wikipedia's Flesch-Kincaid page.

Reading EaseRough US school levelHow it reads
90 to 1005th gradeVery easy
80 to 906th gradeEasy, conversational
70 to 807th gradeFairly easy
60 to 708th and 9th gradePlain English
50 to 6010th to 12th gradeFairly difficult
30 to 50CollegeDifficult
10 to 30College graduateVery difficult
0 to 10ProfessionalExtremely difficult

The scale isn't truly capped at 0 and 100. A passage of one-syllable, one-word sentences can score above 100, and dense academic prose can go negative. Most calculators just show the raw result.

The Flesch-Kincaid Grade Level Formula, Worked by Hand

J. Peter Kincaid and colleagues built the grade level version for the US Navy. Their 1975 report, Derivation of New Readability Formulas for Navy Enlisted Personnel, recalculated older formulas against reading tests taken by 531 Navy personnel. Strictly, only this grade formula carries Kincaid's name, though most tools report both scores as the Flesch-Kincaid readability tests.

Grade Level = (0.39 x ASL) + (11.8 x ASW) - 15.59

Here's a short passage from a typical README, scored by hand:

Run the install script before you start. It creates a config file in your home folder. Edit that file to set your default editor.

The counts come out like this:

SentenceWordsSyllables
Run the install script before you start.79
It creates a config file in your home folder.912
Edit that file to set your default editor.812
Total2433

So ASL is 24 / 3 = 8.0, and ASW is 33 / 24 = 1.375. Now plug them in:

Reading Ease = 206.835 - (1.015 x 8.0) - (84.6 x 1.375)
             = 206.835 - 8.12 - 116.325
             = 82.39

Grade Level  = (0.39 x 8.0) + (11.8 x 1.375) - 15.59
             = 3.12 + 16.225 - 15.59
             = 3.76

The passage scores about 82 for Reading Ease (easy) and about grade 3.8. Your own counts may differ by a syllable or two. "Creates" is two syllables to a person, but a vowel-group counter may disagree, and that's normal.

Is a High Flesch-Kincaid Score Good or Bad?

It depends which Flesch-Kincaid readability score you mean. A high Reading Ease is good if you want broad readability, because it means short sentences and short words. A high Grade Level means the opposite: the text needs more years of schooling to follow.

For everyday documents, Microsoft's readability help page suggests aiming for a Reading Ease of 60 to 70 and a grade level of about 7 to 8. Treat those as editorial conventions, not rules. Nobody tested your audience.

A higher grade is fine in some places. API references, legal terms and research papers use long technical words that you can't swap out. In those cases, watch sentence length and accept a grade of 11 or 12.

We prefer treating the grade level as a smoke alarm for long sentences, not a target to hit. If a page jumps three grades after an edit, look for the sentence that grew. Don't rewrite the whole page to shave off half a point.

Gunning Fog, SMOG, Coleman-Liau and the Fry Readability Formula

Flesch isn't the only option. Readability formulas differ mainly in what they count, and that choice decides how stable they are across tools.

Gunning Fog (Robert Gunning, 1952) counts "complex" words of three or more syllables. The usual counting rules leave out proper nouns, familiar jargon and compound words, and don't treat suffixes like -ed and -ing as an extra syllable.

Fog = 0.4 x (ASL + 100 x complex words / words)

SMOG (G. Harry McLaughlin, 1969) counts polysyllables, again words of three or more syllables. It was normed on 30-sentence samples, so scores for short texts are rough at best.

SMOG = 1.0430 x sqrt(polysyllables x 30 / sentences) + 3.1291

Coleman-Liau (Meri Coleman and T. L. Liau, 1975) counts letters instead of syllables. L is letters per 100 words and S is sentences per 100 words.

Coleman-Liau = (0.0588 x L) - (0.296 x S) - 15.8

The Fry readability formula (Edward Fry, 1968) isn't an equation. You take 100-word samples, count sentences and syllables in each, average them, and plot the point on Fry's graph. The region where the point lands gives the grade. Because it needs a chart lookup, many calculators skip it, and our checker doesn't report Fry.

Automated Readability Index: Counting Characters, Not Syllables

The automated readability index (ARI) comes from a 1967 US Air Force research report by R. J. Senter and E. A. Smith. It was designed for machines: the goal was real-time readability monitoring on electric typewriters, where letters and spaces are easy to count and syllables aren't.

ARI = 4.71 x (characters / words) + 0.5 x (words / sentences) - 21.43

Characters here means letters and digits, not spaces or punctuation. The output is a US grade level. The original method rounds any fraction up to the next whole grade, though many modern tools show the decimal.

For our README passage, 103 letters across 24 words gives 4.29 characters per word:

ARI = (4.71 x 4.29) + (0.5 x 8.0) - 21.43
    = 20.21 + 4.0 - 21.43
    = 2.78

ARI's big advantage is consistency. Counting characters is exact, while counting syllables is a guess. Two tools that split sentences and count characters the same way should report the same ARI, even when their Flesch scores disagree.

Here's the same 24-word README passage run through all six formulas:

FormulaCountsOutputOur passage
Flesch Reading EaseSentence length, syllables0 to 100 score82.4
Flesch-Kincaid GradeSentence length, syllablesUS grade3.8
Gunning FogSentence length, 3+ syllable wordsUS grade4.9
SMOG3+ syllable words per 30 sentencesUS grade6.4 (too few sentences)
Coleman-LiauLetters, sentences per 100 wordsUS grade5.7
ARICharacters, sentence lengthUS grade2.8

One passage, grades from 2.8 to 6.4. That spread is normal, and it's why you should compare a draft against itself over time, using one formula.

How to Check a Readability Score in Word, Grammarly and Hemingway

Word, Grammarly and Hemingway all show some form of readability score, but they don't present the same numbers.

Microsoft Word shows Flesch Reading Ease and Flesch-Kincaid Grade Level. Turn them on first, per Microsoft's help page:

  • Windows: go to File > Options > Proofing and select Check grammar with spelling and Show readability statistics.
  • Mac: go to Word > Preferences > Spelling & Grammar and select the same two options.

Then run a spelling and grammar check. The statistics appear when it finishes.

In Microsoft 365 on the Current Channel, you can instead open Home > Editor > Document stats.

Grammarly lists a readability score among its document statistics, next to word count and vocabulary. As of November 2026, its help center says you open a document in the Grammarly Editor and click Overall score at the top to see them. It doesn't say how the Grammarly readability score is calculated.

Hemingway Editor shows a grade level for your text. Its help page explains the idea behind grade levels but, when we checked, didn't name the formula behind its number.

So why do readability checkers give different scores for the same text? Three reasons account for most of it:

  • Syllable counting. Each tool uses its own heuristic, so words like "creates", "every" or "README" get different counts.
  • Sentence splitting. Abbreviations like "e.g.", version numbers like "v2.1" and list items without a full stop change the sentence count.
  • The formula itself. One app may show Flesch-Kincaid while another shows a different grade formula under the same "grade level" label.

Running a Readability Test on Markdown Docs

Markdown files fool a naive readability test. The formulas expect prose, but a README is full of headings, code and links. Here's a typical snippet:

## Install

- Node 20 or later
- A terminal

Run `npm install` then open http://localhost:3000 in your browser.

A raw count sees no full stop until the very end. "Install Node 20 or later A terminal Run npm install then open..." becomes one long sentence. The URL turns into one enormous "word" with lots of characters, which drags ARI and Coleman-Liau up.

Before you score a doc, take out the parts that aren't prose:

  • Fenced code blocks and inline code. Commands aren't sentences.
  • URLs and image paths. Keep the link text, drop the address.
  • Front matter. YAML keys at the top of a file aren't sentences either.
  • Headings and list items. Either remove them or add a full stop to each, so they count as short sentences.

Then score each prose section on its own. A tutorial with an easy intro and a dense configuration section averages to a misleading middle number.

For structure, our Markdown documentation guide covers how to lay out docs, and the Markdown headings guide covers heading syntax. Short, descriptive headings help readers scan, even though no formula can measure that.

Edits That Improve a Readability Score

Every formula above rewards the same two things: shorter sentences and shorter words. Here's a sentence pair before editing:

Before initiating the installation, execute the setup script, which generates a configuration file in your home directory. You can subsequently modify this configuration file to specify your preferred default editor.

That's 30 words, 2 sentences and 65 syllables. It scores a Reading Ease of about 8 and a grade of about 15.8. The rewrite is the README passage we scored earlier: Reading Ease 82.4, grade 3.8, with the same instructions.

The edits that made the difference:

  • Split long sentences. Anything over about 20 words is a candidate. Cut at "which", "and" or a semicolon.
  • Swap long words for short ones. "Initiating" becomes "start", "execute" becomes "run", "subsequently" becomes "then".
  • Turn nouns back into verbs. "Perform an installation" becomes "install".
  • Use active voice. "The file is created by the script" becomes "The script creates the file". Our passive voice checker guide shows how to find passive sentences in Grammarly, Word and Hemingway.
  • Put sequences in a numbered list. Each step becomes its own short sentence.

Keep the technical terms your readers search for. "Configuration file" is three syllables longer than "config file", but if your audience types the long form, use it and shorten the sentence around it instead. Our Markdown style guide helps keep those word choices consistent across a project.

Here's the limitation every readability formula shares: it measures length, not meaning. A short sentence full of unexplained jargon scores well and still confuses. Syllable counting is also an English heuristic, so scores for other languages aren't meaningful. Use the number to find long sentences, then read the result aloud.

A Flesch-Kincaid readability calculator is most useful as a before-and-after check. Score a draft, split the longest sentences, swap the heaviest words, then score it again with the same tool. Strip code, URLs and front matter from Markdown first so the counts reflect your prose. Paste your next draft into our free readability checker, or write and revise it in the Markdown editor, which shows a live word count as you type.