Readability formulas estimate how hard a text is from two counts: how long the sentences are and how long the words are. Flesch Reading Ease turns them into a score from 0 to 100, where higher means easier. Flesch-Kincaid, Gunning Fog, SMOG, Coleman-Liau and the Automated Readability Index turn them into a U.S. school grade. They rarely agree, because each formula weighs different counts and every tool counts syllables and sentences its own way. Read the results as a range, not as an exact grade.
This post shows what each formula measures, works through one paragraph by hand, and explains the four reasons two tools give different scores for the same text. To follow along, paste the examples into the FindUtils Readability Calculator, which shows all six scores at once.
What the six formulas count
Every formula uses average sentence length, words divided by sentences. They differ in how they measure word difficulty:
| Formula | Word difficulty measured by | Result | Formula used by most calculators |
|---|---|---|---|
| Flesch Reading Ease | Syllables per word | 0 to 100, higher is easier | 206.835 − 1.015 × ASL − 84.6 × ASW |
| Flesch-Kincaid Grade | Syllables per word | U.S. grade | 0.39 × ASL + 11.8 × ASW − 15.59 |
| Gunning Fog | Share of words with 3+ syllables | U.S. grade | 0.4 × (ASL + 100 × complex words ÷ words) |
| SMOG | Count of words with 3+ syllables | U.S. grade | 1.043 × √(complex words × 30 ÷ sentences) + 3.1291 |
| Coleman-Liau | Characters per 100 words | U.S. grade | 0.0588 × L − 0.296 × S − 15.8 |
| Automated Readability Index | Characters per word | U.S. grade | 4.71 × characters ÷ words + 0.5 × ASL − 21.43 |
ASL is the average sentence length in words, ASW the average syllables per word. In Coleman-Liau, L is characters per 100 words and S is sentences per 100 words.
The formulas come from different decades and different readers. Rudolf Flesch published Reading Ease in 1948 (Journal of Applied Psychology). Peter Kincaid and colleagues recalculated it as a grade level in 1975 for the U.S. Navy, testing 531 enlisted personnel on passages from training manuals. Harry McLaughlin designed SMOG in 1969 to be counted by hand in 30 sentences. Coleman-Liau (1975) and the Automated Readability Index (1967) were built for machines and avoid syllables altogether: ARI was first counted from the keystrokes of a typewriter.
One paragraph, worked by hand
Here is a 70-word paragraph:
Readability formulas count what is easy to count. They look at how long your sentences are and how long your words are. They do not know whether your reader cares about the topic, knows the terms, or can follow the order of your ideas. A short sentence full of unfamiliar words can score well and still confuse people. Use the score as a quick check, then ask a real reader.
The Readability Calculator counts 70 words, 5 sentences and 92 syllables. So ASL = 70 ÷ 5 = 14 and ASW = 92 ÷ 70 ≈ 1.314:
- Flesch Reading Ease = 206.835 − 1.015 × 14 − 84.6 × 1.314 ≈ 81.4 (Easy)
- Flesch-Kincaid Grade = 0.39 × 14 + 11.8 × 1.314 − 15.59 ≈ 5.4
All the grade-level scores from the same paste:
| Score | Result |
|---|---|
| Flesch-Kincaid Grade | 5.4 |
| Automated Readability Index | 6.1 |
| Gunning Fog | 7.3 |
| SMOG | 7.6 |
| Coleman-Liau | 7.7 |
| Average grade level | 6.8 |
Same text, same tool, and the grade levels run from 5.4 to 7.7. That is normal. The CMS Toolkit for Making Written Material Clear and Effective warns that scores for the same text "can differ by two, three, or more grade levels" depending on the formula and how you use it, and that Flesch-Kincaid tends to come out lower than other measures.
Why two tools give different scores
Even with the same formula, two checkers often disagree. Four things cause almost all of it.
1. Syllables are guessed, not looked up
Software cannot hear a word, so it estimates syllables. Some tools look words up in a pronouncing dictionary such as the CMU Pronouncing Dictionary. Others, including the Readability Calculator, apply spelling rules: count the vowel groups, then drop a silent final e. Rules are fast and work on any word, but English spelling fights back in both directions:
| Word | Spelling rules (Readability Calculator) | Pronouncing dictionary |
|---|---|---|
| every | 3 | 2 |
| business | 3 | 2 |
| different | 3 | 2 |
| created | 1 | 3 |
| idea | 2 | 3 |
| science | 1 | 2 |
| sentences | 2 | 3 |
In the paragraph above, the dictionary finds 94 syllables instead of 92. That moves Flesch-Kincaid from 5.4 to 5.7. The bigger effect is on the "complex word" count that Fog and SMOG use: the dictionary finds 5 words with three or more syllables instead of 3, which lifts Gunning Fog from 7.3 to 8.5 and SMOG from 7.6 to 8.8. One or two words decide a whole grade.
2. Sentences end at every full stop
Most checkers end a sentence at each period, question mark or exclamation mark, the same rule McLaughlin gave for counting SMOG by hand. Abbreviations and decimals break it. Add "e.g. before you publish" to the last sentence of the paragraph, and the calculator counts 7 sentences instead of 5. Flesch-Kincaid drops from 5.4 to 4.2, and the text did not get any easier.
Headings and bullet points do the opposite. With no full stop at the end, a heading is glued onto the next sentence. Put the line "How readability scores work" above the paragraph and Flesch-Kincaid rises from 5.4 to 6.1. The CMS toolkit gives the same advice for any computer scoring: remove periods used in abbreviations, and remove titles, headings and bullet points that are not full sentences before you score.
3. "Characters" means different things
Coleman-Liau and ARI count characters instead of syllables, but tools disagree on which characters. The Readability Calculator counts every character that is not a space, punctuation included: 305 for the paragraph. Count letters only and you get 297, which lowers Coleman-Liau from 7.7 to 7.0 and ARI from 6.1 to 5.6.
4. Short texts and limits
SMOG was designed for a 30-sentence sample. The calculator scales the long-word count to a 30-sentence equivalent, so in a five-sentence paragraph every long word counts six times. Flesch Reading Ease is capped at 0 and 100 on the calculator, while the raw formula can go past either end. The calculator needs at least 10 words to score anything; aim for 100 or more so a single sentence does not swing the result.
What the numbers mean
The Readability Calculator labels Flesch Reading Ease with these bands:
| Flesch Reading Ease | Label |
|---|---|
| 90 to 100 | Very Easy |
| 80 to 89 | Easy |
| 70 to 79 | Fairly Easy |
| 60 to 69 | Standard |
| 50 to 59 | Fairly Difficult |
| 30 to 49 | Difficult |
| 0 to 29 | Very Difficult |
Grade levels map to audiences: up to grade 5 is elementary school, up to 8 middle school, up to 12 high school, up to 16 college, and above that graduate or professional. Microsoft's Word readability help suggests a Flesch Reading Ease between 60 and 70 for most standard documents, and a Flesch-Kincaid grade of about 7 to 8.
How to get a score you can trust
- Paste prose only. Remove headings, bullet lists, captions and code, or end them with a full stop if you want them counted as sentences.
- Expand abbreviations such as "e.g.", "i.e." and "Dr." before you score, or accept that they make the text look easier.
- Use a big enough sample. At least 100 words; for a long document, take passages from the beginning, middle and end, as SMOG's instructions do.
- Compare like with like. Score your before and after versions in the same tool. A score from one checker is not comparable to a score from another.
- Look at the spread, not one number. If Flesch-Kincaid says 6 and SMOG says 9, the honest answer is "somewhere between 6 and 9".
To see where the length is coming from, the Word Counter shows sentence and paragraph counts, and the Reading Time Estimator tells you how long the page takes to read.
Do not write for the formula
A formula only sees length. It cannot tell whether your reader knows the terms, whether the steps are in a sensible order, or whether a short sentence is short because a needed word was cut. The CMS toolkit recommends using readability formulas as a check on words and sentences, not as a measure of comprehension or a guide to writing, and warns that chasing a grade level can produce choppy text that lacks cohesion. Shorten a sentence when it is hard to follow, swap a long word when a shorter familiar one exists, and then test the text with a real reader.
Key takeaways
- Every readability formula is built from sentence length plus word length, measured in syllables or characters.
- Flesch Reading Ease runs from 0 to 100 (higher is easier); the other five give a U.S. grade level.
- The same text can score two or more grades apart across formulas, and across tools using the same formula.
- Syllable guessing, sentence splitting at abbreviations and headings, and character definitions explain most disagreements.
- Clean the text, score at least 100 words, compare versions in the same tool, and treat the result as a range.
FAQ
What is a good Flesch Reading Ease score?
For most general writing, 60 to 70 is a common target; Microsoft's Word help recommends that range for standard documents. Above 80 is easy to read, and below 50 is hard for a general audience.
What is the difference between Flesch Reading Ease and Flesch-Kincaid?
They use the same two counts, words per sentence and syllables per word, with different weights. Reading Ease gives a 0 to 100 score where higher is easier. Flesch-Kincaid converts the counts into a U.S. school grade, where lower is easier.
Why does my readability score differ between tools?
Tools count syllables differently (spelling rules or a dictionary), split sentences differently (abbreviations, headings and lists), and define characters differently. Each of those can move a grade level by half a grade or more on a short text.
Which readability formula is most accurate?
None is accurate for every text. SMOG and Gunning Fog focus on long words, Coleman-Liau and ARI on characters, and the two Flesch formulas on syllables. Use several, read the range, and test with real readers when it matters.
How many words do I need for a reliable readability score?
At least 100. The Readability Calculator needs 10 words to produce scores, but with very short texts one sentence or one long word changes the result a lot. SMOG was designed for 30 sentences.
Is there a free readability checker?
Yes. The FindUtils Readability Calculator is free, needs no signup, and shows Flesch Reading Ease, Flesch-Kincaid, Gunning Fog, SMOG, Coleman-Liau, ARI and an average grade level at once, with the text processed in your browser.
Related reading
- Word counter guide: words, characters, sentences and reading time
- Find and replace text online, including regex
- Case converter guide: title case, sentence case and more
- Why 10 sorts before 2: how text line sorting works
Need help with a project? The team behind FindUtils takes on freelance work, from a single bug fix to an AI feature in your web app such as document analysis.
Sources
- Flesch, R. (1948). A new readability yardstick. Journal of Applied Psychology, 32, 221–233
- Kincaid, J. P., Fishburne, R. P., Rogers, R. L. and Chissom, B. S. (1975). Derivation of New Readability Formulas for Navy Enlisted Personnel
- McLaughlin, G. H. (1969). SMOG Grading: a New Readability Formula. Journal of Reading
- Coleman, M. and Liau, T. L. (1975). A computer readability formula designed for machine scoring. Journal of Applied Psychology, 60, 283–284
- Smith, E. A. and Senter, R. J. (1967). Automated Readability Index (PubMed record)
- U.S. Centers for Medicare & Medicaid Services: Toolkit for Making Written Material Clear and Effective, Part 7, Using readability formulas: a cautionary note (copy hosted by Montefiore Einstein)
- Microsoft Support: Get your document's readability and level statistics
- CMU Pronouncing Dictionary