What Is Flesch Reading Ease? (Formula + How to Improve Your Score)

The Flesch Reading Ease score is the most widely used readability formula in the English language. It's built into Microsoft Word, Google Docs, Grammarly, Yoast SEO, the Hemingway Editor, and pretty much every readability tool you've ever used. It's the score that quietly runs in the background when WordPress tells you "your content is easy to read."

But what does the score actually measure? Where did the formula come from? And — most importantly — if your score is low, what specific things can you do to bring it up without dumbing down your content?

This article walks through the formula, works a real example with real numbers, explains what the score ranges actually mean, and gives you concrete, non-generic techniques for improving your score.

Where the Formula Came From

The Flesch Reading Ease formula was published in 1948 by Rudolf Flesch, an Austrian-born writer and readability researcher who had emigrated to the United States in the 1930s. Flesch developed it while working on his Columbia University doctoral dissertation, Marks of a Readable Style, which later became his 1946 book The Art of Plain Talk.

Flesch's goal was to create a formula that predicted how difficult a piece of English prose would be to read, based only on measurable surface features of the text. He was building on earlier 20th-century readability research — most notably the 1935 study by William S. Gray and Bernice Leary, What Makes a Book Readable, which had identified word difficulty and sentence length as the two strongest predictors of readability through a regression analysis of some 300 passages and 1,600 adult readers. Flesch refined their methodology into a simpler arithmetic formula that could be computed by hand from any text.

The formula Flesch published in 1948, in the Journal of Applied Psychology, was:

Reading Ease = 206.835 − (1.015 × ASL) − (84.6 × ASW)

Where:

ASL = Average Sentence Length (words per sentence)
ASW = Average Syllables per Word

The constants (206.835, 1.015, 84.6) were fitted by regression against Flesch's dataset of reading-comprehension test results from school-age readers. The output is a number on a 0–100 scale, where higher scores indicate easier-to-read text.

This is the same formula in use today. It hasn't been updated or revised in 75 years. That's both its strength (it's well-validated) and its weakness (it was calibrated on 1940s reading tests and 1940s prose).

The 0–100 Scale and What It Means

Flesch's original interpretation of the score ranges, still in use:

| Score | Reading level | Example |

|---|---|---|

| 90–100 | 5th grade — very easy to read | Comic strips, marketing copy for kids |

| 80–90 | 6th grade — easy | Consumer marketing copy, popular magazines |

| 70–80 | 7th grade — fairly easy | Reader's Digest, typical blog posts |

| 60–70 | 8th–9th grade — plain English | Most business writing, news articles |

| 50–60 | 10th–12th grade — fairly difficult | Time, Newsweek; quality newspapers |

| 30–50 | College — difficult | Academic journals, professional magazines |

| 0–30 | College graduate — very difficult | Legal documents, academic papers, scientific papers |

Some famous reference points:

Reader's Digest averages around 65.
Time magazine averages around 52.
The New York Times averages around 50 (the front page; opinion pieces run lower).
Harvard Law Review averages around 30.
Internal Revenue Code averages in the low 20s.
A typical Dr. Seuss book scores 100+ (very short sentences, mostly one-syllable words).

For comparison, the average U.S. adult reads at about an 8th-grade level. The National Assessment of Adult Literacy (2003, the most recent comprehensive study) found that 43% of U.S. adults read at or below a "basic" level — roughly 8th grade. Most public-facing writing guidelines therefore recommend targeting a Flesch score of 60 or higher for general audiences.

The Formula, Step by Step

Let's work through the formula with real text. Here's a sample paragraph we'll score:

The implementation of comprehensive regulatory frameworks governing artificial intelligence systems requires careful consideration of multifaceted ethical implications, including algorithmic transparency, demographic equity, and accountability mechanisms. Recent legislative initiatives in jurisdictions worldwide reflect growing consensus regarding the necessity of such frameworks.

Step 1: Count words

We count 36 words. (We'll need this for both ASL and ASW.)

Step 2: Count sentences

There are 2 sentences. (Sentence boundaries: the period after "mechanisms" and the period after "frameworks.")

Step 3: Calculate ASL (Average Sentence Length)

ASL = Words ÷ Sentences = 36 ÷ 2 = 18.0

Step 4: Count syllables in each word

Going word by word:

The (1), implementation (5), of (1), comprehensive (4), regulatory (5), frameworks (2), governing (3), artificial (4), intelligence (4), systems (2), requires (3), careful (2), consideration (5), of (1), multifaceted (5), ethical (3), implications (4), including (3), algorithmic (4), transparency (4), demographic (4), equity (3), and (1), accountability (6), mechanisms (4). = 82 syllables
Recent (2), legislative (4), initiatives (5), in (1), jurisdictions (4), worldwide (2), reflect (2), growing (2), consensus (3), regarding (3), the (1), necessity (4), of (1), such (1), frameworks (2). = 37 syllables

Total: 82 + 37 = 119 syllables.

Step 5: Calculate ASW (Average Syllables per Word)

ASW = Syllables ÷ Words = 119 ÷ 36 = 3.31

Step 6: Plug into the Flesch formula

Reading Ease = 206.835 − (1.015 × 18.0) − (84.6 × 3.31)
             = 206.835 − 18.27 − 280.03
             = −91.46

That's a negative score, which Flesch's scale clamps to 0. The paragraph is, predictably, very difficult to read — it's full of long, abstract, multi-syllable words and long sentences. A score of 0 (or below) corresponds to "college graduate" reading level, which is accurate: this paragraph reads like a law review article.

Compare with a rewrite

Here's the same idea, rewritten:

Making rules for AI is hard. We have to think about fairness, transparency, and who's responsible when things go wrong. More countries are starting to agree we need these rules.

Word count: 33. Sentences: 3. Syllables: ~45.

ASL = 33 ÷ 3 = 11.0
ASW = 45 ÷ 33 = 1.36
Reading Ease = 206.835 − (1.015 × 11.0) − (84.6 × 1.36)

= 206.835 − 11.17 − 115.06

= 80.6

That's an 80.6 — "easy," 6th grade reading level. Same information, dramatically different readability. The rewrite cut average sentence length roughly in half and average word length by more than half.

This is the whole point of Flesch Reading Ease: it gives you a number that responds to the two variables that matter most for readability — sentence length and word length.

What the Formula Gets Right (and Wrong)

Flesch Reading Ease is the most-used readability formula for good reasons:

It works. Decades of validation studies have shown that the Flesch score correlates with actual reading comprehension at about r = 0.7 — meaning it explains roughly 50% of the variance in comprehension difficulty. For a formula based purely on word and sentence length, that's a strong result.

It's simple. The formula uses only two measurements that any text editor can compute automatically. No need for vocabulary lists, syntactic parsing, or domain expertise.

It's interpretable. The 0–100 scale maps roughly to U.S. grade levels, which most adults can reason about. A score of 60 means "8th grade" — concrete and actionable.

But the formula has real limitations:

Limitation 1: It doesn't account for word difficulty directly

A long word isn't necessarily a hard word. "Implementation" is five syllables but well-known to most adults. "Syzygy" is three syllables and obscure. Flesch penalizes both the same way. A text that uses long technical jargon that readers actually know (because they're domain experts) scores artificially low.

Limitation 2: It doesn't account for sentence structure

A 20-word sentence with a simple subject-verb-object structure is easier to read than a 20-word sentence with three embedded clauses and a parenthetical. Flesch can't tell the difference.

Limitation 3: It doesn't account for cohesion

A paragraph where each sentence builds on the last is easier to read than a paragraph where sentences jump between topics. Flesch measures nothing about cohesion, transition words, or logical flow.

Limitation 4: It doesn't account for formatting

A paragraph broken into 5 lines with white space reads easier than a 5-line block of dense text. Flesch doesn't see paragraphs, headings, bullet points, or images.

Limitation 5: It's calibrated on 1940s prose and 1940s readers

Vocabulary has shifted. Sentence structures in published prose have changed. The 7th-grade reading level of 1948 isn't quite the same as the 7th-grade reading level of 2025. The formula's constants are slightly stale, but no one has produced a better-validated replacement.

For all these limitations, the formula is still useful because sentence length and word length really are the strongest predictors of readability. The things it misses matter, but they matter less than the two things it captures.

What's a Good Flesch Reading Ease Score?

It depends on your audience and content type. Here's a practical guide:

| Content type | Target Flesch score | Reading level |

|---|---|---|

| Children's content (under 12) | 90–100 | 5th grade |

| Marketing copy for general consumers | 70–80 | 7th grade |

| Blog posts for general audiences | 60–70 | 8th–9th grade |

| News articles | 50–60 | 10th grade |

| B2B blog posts and white papers | 40–50 | 11th–12th grade |

| Academic papers | 30 or below | College |

| Legal documents | 30 or below | College graduate |

| Government public-facing content | 60+ | Per Plain Writing Act of 2010 |

| Medical information for patients | 80+ | 6th grade (per AMA guidance) |

| Technical documentation | 50–60 | 10th grade |

The U.S. Plain Writing Act of 2010 actually requires federal agencies to write at a level the public can understand — operationalized as a Flesch score of 60 or higher. The American Medical Association recommends patient-facing materials be written at a 6th-grade level (Flesch 70–80). These are legal or professional standards, not just preferences.

For most web content aimed at general audiences, a Flesch score of 60 or higher is a sensible target. Below 50, you're losing readers. Below 30, you're losing almost everyone who isn't a specialist.

How to Improve Your Flesch Reading Ease Score

If your score is lower than your target, here are the specific, non-generic techniques that actually move the number. Each one works because it changes one of the two variables in the formula.

1. Cut sentence length in half

The fastest way to raise your score. A 25-word sentence split into two 12-word sentences raises your ASL by 50%, which has an outsized effect on the score.

Before (one 26-word sentence, Flesch ≈ 45):

Despite the fact that the implementation timeline was accelerated by the project sponsor, the development team was able to deliver all critical features on schedule and within the approved budget allocation.

After (three sentences, Flesch ≈ 65):

The project sponsor accelerated the implementation timeline. Even so, the development team delivered all critical features on schedule. They stayed within the approved budget.

Same information, ~20 words shorter, much higher score.

2. Replace Latinate words with Germanic equivalents

English has two main vocabulary layers: Germanic (short, common, one-syllable) and Latinate (long, formal, multi-syllable). Replacing Latinate words with their Germanic equivalents is one of the highest-leverage moves for readability.

| Latinate (lower score) | Germanic (higher score) |

|---|---|

| Utilize | Use |

| Implement | Do |

| Terminate | End |

| Initiate | Start |

| Endeavor | Try |

| Facilitate | Help |

| Demonstrate | Show |

| Obtain | Get |

| Require | Need |

| Sufficient | Enough |

| Approximately | About |

| Numerous | Many |

| Subsequently | Then |

| Currently | Now |

| Subsequently | Later |

None of these swaps change meaning. They cut one to three syllables per word, which directly raises your ASW and your score.

3. Replace abstract nouns with their verb forms

Nominalization — turning verbs into nouns — is a major cause of low Flesch scores. "Make a decision" (3 words, 5 syllables) vs. "decide" (1 word, 2 syllables). "Conduct an investigation" (3 words, 7 syllables) vs. "investigate" (1 word, 4 syllables).

Before: "We will conduct an investigation into the matter."

After: "We will investigate the matter."

Three words instead of eight. Score climbs.

4. Cut filler phrases

Common filler that adds words without meaning:

"In order to" → "to" (saves 2 words)
"Due to the fact that" → "because" (saves 5 words)
"At this point in time" → "now" (saves 4 words)
"In the event that" → "if" (saves 3 words)
"For the purpose of" → "for" (saves 3 words)
"With regard to" → "about" (saves 2 words)
"It should be noted that" → cut entirely (saves 5 words)
"There is/there are" → restructure (saves 2 words)

Each filler phrase adds a sentence-length penalty. Cut them.

5. Use contractions where natural

"Cannot" → "can't." "Will not" → "won't." "It is" → "it's." Each contraction cuts one syllable per instance, and the resulting word is shorter, which improves ASW slightly. More importantly, contractions match how people actually speak, which improves perceived readability.

6. Replace passive voice with active

Passive voice constructions are longer and more abstract, on average, than active equivalents. "The report was reviewed by the committee" (7 words, 11 syllables) vs. "The committee reviewed the report" (5 words, 9 syllables). Less direct, longer, and less clear.

7. Replace multi-word phrases with single words

"A large number of" → "many"
"A small number of" → "few"
"In the near future" → "soon"
"On a regular basis" → "regularly"
"In the process of" → cut
"With the exception of" → "except"

8. Break up paragraphs

While Flesch doesn't measure paragraph length directly, breaking long paragraphs into shorter ones forces shorter sentences (because each paragraph tends to be 1–3 sentences). The visual effect also makes content feel easier to read, which affects perceived readability.

9. Replace jargon with plain language

Domain-specific jargon is usually longer and less familiar than plain equivalents. "Cognitive dissonance" (6 syllables) vs. "mental conflict" (4 syllables). "Paradigm shift" (3 syllables) vs. "big change" (2 syllables). If your audience isn't composed of specialists, replace jargon.

10. Read your draft aloud

This is the lowest-tech readability check. Anywhere you stumble, gasp for breath, or lose your place — that's where readers will struggle too. Cut and rewrite those passages.

A Caveat: Don't Optimize Blindly

Flesch Reading Ease is a useful indicator, not an end in itself. A few cautions:

Don't shorten sentences by creating choppy, disconnected ones. Five short sentences in a row can be harder to read than one well-structured long sentence, because cohesion suffers.
Don't replace precise words with vague ones. "Use" and "utilize" sometimes mean the same thing; sometimes "utilize" specifically means "use for a new purpose." If the precision matters, keep the longer word.
Don't dumb down technical content. If you're writing for cardiologists, a Flesch score of 35 is appropriate. If you dumb it down to 70, you'll lose the audience.
Don't optimize at the expense of voice. Some writers' natural style produces lower scores because they use long, sonorous words. If your voice works, don't homogenize it for a number.

The score is a diagnostic, not a goal. Aim for the target range for your content type, but trust your judgment about what reads well.

Tools That Compute Flesch Reading Ease

If you want to check your score:

Microsoft Word: File → Options → Proofing → "Show readability statistics." Runs after spell-check.
Google Docs: Add-ons like ProWritingAid or Readable.
Yoast SEO / Rank Math: WordPress plugins that show Flesch score in the editor.
Hemingway Editor: Web-based tool that shows reading grade level (closely related to Flesch).
Our readability analyzer: Runs Flesch Reading Ease, Flesch-Kincaid Grade Level, Gunning Fog, and other formulas on your text in real time. No upload or signup required.

The Bottom Line

Flesch Reading Ease has been the dominant readability formula for 75 years because it captures the two biggest predictors of reading difficulty: sentence length and word length. It's not perfect — it can't measure cohesion, structure, or familiarity — but it's a reliable first-pass signal.

For general web content, aim for 60 or higher. For consumer-facing copy, aim for 70+. To raise your score, shorten your sentences (the single biggest lever), replace Latinate words with Germanic equivalents, and cut filler phrases. The number will move.

But remember: the score is a diagnostic, not the goal. A Flesch score of 80 on content that doesn't say anything is worse than a score of 50 on content that does. Use the formula to identify problems, then use your judgment to fix them.

Our readability tool gives you a live Flesch Reading Ease score as you type, along with the Flesch-Kincaid Grade Level, sentence-length breakdowns, and syllable counts.