Readability Checker

What the score measures — the actual math

The Flesch Reading Ease score, published by Rudolf Flesch in 1948, reduces readability to two measurable habits: how long your sentences run, and how long your words are. The formula is:

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

Higher is easier, on a 0–100 scale. The companion Flesch-Kincaid Grade Level — derived in 1975 for the US Navy, which needed to know whether recruits could read their technical manuals — converts the same two inputs into school grades:

0.39 × (words ÷ sentences) + 11.8 × (syllables ÷ words) − 15.59

Notice what the formulas reward. Nothing about vocabulary difficulty, logic, or structure — only sentence length and syllable load. That makes the score a blunt instrument, but a useful one: long sentences stuffed with long words are, in fact, the most common way drafts go wrong. This tool also grades every sentence individually and surfaces the three hardest, because a mediocre average often hides two paragraphs of trouble in an otherwise clean draft.

Reading the score

For calibration: Reader’s Digest historically scored around 65, Time around 52, and the Harvard Law Review in the low 30s. All three are professionally edited; they simply serve different readers. That is the right way to hold this number — as a match to an audience, not a grade on your intelligence.

What to aim for

For blog posts, landing pages, documentation, and email to a general audience: a Reading Ease of 60–70, a grade level of 8–9 or below, sentences averaging 15–20 words, and about 1.4–1.6 syllables per word. Those four numbers describe most successful web writing. Instructions and health or safety content should go simpler still — US federal plain-language guidance effectively targets grade 6–8. Writing for specialists can sit lower on the scale without apology; “myocardial infarction” is the correct term in a cardiology journal, and no formula should talk you out of it.

One number the formula hides: variance. An average of 17 words per sentence can describe lively prose that alternates 8-word and 26-word sentences, or a monotonous march of 17-word sentences that reads like a metronome. Professional prose keeps the average in range while mixing lengths deliberately — a short sentence lands a point, a longer one develops it. So treat the average as a boundary condition and the hardest-sentences list as the real work order: if the three flagged sentences all sit above grade 15 while the draft averages grade 8, the problem is two specific passages, not your style.

How to use it well

1. Paste the finished draft — at least a few paragraphs. Scores on fragments are noise.
2. Read the number against the table above and ask one question: does this match the reader I actually have?
3. Go straight to the hardest sentences the tool flags and read them aloud. If you run out of breath or lose the thread, so will the reader.
4. Fix by splitting first, substituting second: break any sentence over 25 words at its conjunctions, then swap the multi-syllable latinate words for short ones.
5. Re-check and stop when you land in range. Grinding from 63 to 71 is polishing a number, not the prose.

Common mistakes

• Chasing the maximum. A 95 built from staccato four-word sentences reads like a children’s primer and erodes trust with adult readers.
• Editing the number instead of the meaning. Mechanically deleting commas and splitting every sentence raises the score while making the logic harder to follow.
• Applying one target to every document. A privacy policy, a product tour, and a research summary have different readers and deserve different bands.
• Scoring text the formula was never built for. Code samples, tables, headings, and bullet fragments distort the sentence math — score the prose, not the page furniture.
• Ignoring structure entirely. The formula cannot see headings, lists, or paragraph breaks, and those often do more for perceived clarity than two grade levels of word choice.

When the score misleads you

Flesch arithmetic knows nothing about meaning. A scrambled sentence scores exactly the same as the original; confident nonsense can score 75 while a precise, necessary technical explanation scores 35. Syllable counting itself is heuristic — English spelling is irregular enough that any automatic counter, including this one, misses edge cases like “poem” or borrowed words, so treat the third decimal place as fiction. The formulas were calibrated on English; scores for Arabic or any other language are not meaningful, even though the word and sentence counts still are. And short is not automatically clear: a ten-word garden-path sentence can stall a reader longer than a well-built thirty-word one. Use the score as a smoke detector — very good at telling you something is burning, useless at telling you what to cook.