Readability Checker
How hard your writing is to read. And, more usefully, which specific sentences are dragging it down.
In short
Flesch Reading Ease combines average sentence length and syllables per word. Aim for 60–70 for a general audience. The fastest way to improve a score is to split the few longest sentences, not to shorten every word.
Last reviewed
Runs entirely in your browser. Nothing you type here is uploaded, stored or logged.
What the score measures
Flesch Reading Ease uses two inputs and nothing else:
206.835 − 1.015 × (words ÷ sentences) − 84.6 × (syllables ÷ words) Long sentences push it down. Long words push it down harder. That's the whole model, which is both its strength, it's objective and instant, and its limitation.
Reading the number
| Score | Level | Typical of |
|---|---|---|
| 90–100 | Very easy | Children's books |
| 80–90 | Easy | Conversational writing |
| 70–80 | Fairly easy | Popular fiction, good plain English |
| 60–70 | Standard | Newspapers, most web writing |
| 50–60 | Fairly difficult | Quality journalism, trade press |
| 30–50 | Difficult | Academic writing, technical documentation |
| Below 30 | Very difficult | Legal contracts, research papers |
60–70 is the sensible default for anything aimed at a general audience. That's roughly an eighth-grade reading level, and it is not a low bar, it's the level at which people read comfortably rather than effortfully, including people who are tired, distracted, reading on a phone, or reading in a second language.
A low score is not automatically bad
A contract, a research paper or a technical specification should score low. Precision costs readability, and in those contexts precision wins. Readers of a journal article expect and can handle dense prose.
The score becomes a problem when it's lower than your audience expects. A privacy policy scoring 25 is a real failure, because the people it exists to inform can't read it. A landing page scoring 40 is losing readers who had a choice.
How to actually raise it
Split long sentences. This is the highest-leverage change by a wide margin. Any sentence over about 25 words is usually two sentences that haven't been separated yet. Look for "and", "but" and semicolons as the natural break points.
Choose shorter words where they mean the same thing. "Use" for "utilise", "help" for "facilitate", "about" for "approximately", "buy" for "purchase", "start" for "commence". Not because long words are wrong, but because the short one is usually clearer and there's no cost.
Cut throat-clearing. "It is important to note that" is six words that carry no information. "In order to" is "to".
Prefer verbs to noun phrases. "We made a decision" → "we decided". "Give consideration to" → "consider". This shortens sentences and makes them more direct at the same time.
Use the active voice by default. "The committee approved the plan" is shorter and clearer than "the plan was approved by the committee". The passive has real uses, when the actor is unknown or irrelevant, but it's usually a habit rather than a choice.
What the score can't see
This is the important caveat. Flesch measures sentence and word length, so it can be gamed and it can be wrong:
- Short and meaningless scores well. "The thing did the thing to the thing" is very readable and says nothing.
- Jargon in short words is invisible to it. "We need to align on the ask before we socialise the deck" scores beautifully.
- It ignores structure. Headings, paragraphs, lists and ordering affect comprehension enormously and don't appear in the formula.
- It ignores whether you're right. Obviously, but worth saying.
Use it as a smoke alarm, not a grade. A bad score reliably means something is wrong. A good score means nothing is obviously wrong with your sentence lengths.
Why the longest sentences are listed
Because that's where the fixable problems concentrate. Improving a document's score is almost always a matter of finding the five worst sentences and splitting them, far more effective than trying to shave a syllable off every word.
Frequently asked questions
What is a good readability score?
For general audiences, aim for Flesch Reading Ease of 60–70, which is roughly an eighth-to-ninth-grade level. Most newspapers land there. Below 30 is genuinely hard going and appropriate only for academic or technical readers who expect it.
Does a low score mean bad writing?
No. It means demanding writing. A legal contract or a research paper should score low; that is the register. The score is only a problem when it is lower than your audience expects, which is why the target matters more than the number.
How is the score calculated?
Flesch Reading Ease combines average sentence length and average syllables per word. Long sentences and long words both push it down. It is a proxy, not a measure of clarity. You can score well and still be incoherent.
How are syllables counted?
By an algorithm that counts vowel groups and adjusts for common patterns like silent E and -le endings. It is accurate for most English words and wrong for some unusual ones. Small errors barely move the score across a whole document.
Does this work for other languages?
Not reliably. Flesch and its relatives were developed for English, and the syllable heuristics assume English spelling. Other languages have their own adaptations with different constants.