Flesch Reading Ease Score
Score
92.1
Level
5th grade
Sentences
2
How it works
The Flesch Reading Ease score is the most widely-used automated readability metric in English. Developed by Rudolph Flesch in 1948, it produces a score between 0 and 100 where higher = easier to read. US government guidelines, US Navy training materials, and many publishing standards specify minimum Flesch scores for different audiences.
**Formula** Flesch Reading Ease = 206.835 − (1.015 × ASL) − (84.6 × ASW) Where: ASL = Average Sentence Length (words per sentence); ASW = Average Syllables per Word.
**Score interpretation** - 90–100: Very easy (5th grade, simple text) - 70–80: Easy (6th grade, conversational prose) - 60–70: Standard (7th–8th grade, most general-audience writing) - 50–60: Fairly difficult (9th–10th grade, academic prose) - 30–50: Difficult (college level, scientific writing) - 0–30: Very difficult (professional/legal documents)
**Flesch-Kincaid Grade Level** The companion formula: FKGL = 0.39 × ASL + 11.8 × ASW − 15.59. This expresses readability as a US school grade level. FKGL 8.0 means an average 8th-grader can read it. Medical informed consent should target ≤8; most US newspapers write at grade 11–14.
**Practical applications** Legal documents (GDPR privacy policies, terms of service, insurance disclosures) are frequently criticised for extreme complexity — Flesch scores of 10–20 are common. The EU's Plain Language initiative targets grade 8 for official communications. Microsoft Word's built-in grammar checker includes Flesch-Kincaid because its importance was recognised in word processing from the start.
**Limitations** Flesch scores work best for English prose. They struggle with code documentation, lists, technical specifications, and non-English languages. A poem may score high (short sentences, short words) while being cognitively complex.
Privacy: all analysis runs in the browser. No text is transmitted.
Frequently Asked Questions
- Legal documents prioritise precision and enforceability over readability — ambiguity has legal consequences. Long sentences contain multiple qualifications and exceptions ('provided, however, that unless otherwise specified herein...'). Technical vocabulary (indemnification, tort, estoppel, subrogation) cannot be simplified without changing legal meaning. Passive voice is preferred to avoid assigning specific obligations. The result: typical insurance policy = 8–15 Flesch score (college professor level). The EU Plain Language initiative and US Plain Writing Act (2010) push back against this, requiring government communications at an 8th-grade reading level.
- Target scores by audience: children's books (ages 8–10): 80–90. Mass-market fiction: 65–75. General-audience magazine articles: 60–70. Business communications and company policies: 50–65. Academic journal papers: 30–50. Technical specifications and manuals: 20–40. Legal contracts: 10–30. Medical informed consent: target 60+ (most consent forms score below 40, which is a public health concern — patients can't give informed consent if they can't understand the document). As a content creator, aim for 60–70 for general audiences unless your readers are specialists.
- In the Flesch formula, Average Sentence Length (ASL) has a coefficient of 1.015 × ASL, while Average Syllables per Word (ASW) has a coefficient of 84.6 × ASW. A sentence 20 words long (vs. 10 words) reduces the score by about 10 points. Using 'utilise' instead of 'use' (3 syllables vs. 1) only reduces the score by 84.6 × (3−1)/word_count per sentence. For a 20-word sentence, replacing one word adds 84.6 × 2/20 = 8.5 points of difficulty. Shorter sentences are the highest-leverage readability improvement.
- Yes. Count: total words (W), total sentences (S), total syllables (Sy). ASL = W/S; ASW = Sy/W. Score = 206.835 − (1.015 × W/S) − (84.6 × Sy/W). For a paragraph of 100 words in 5 sentences (ASL=20) with 145 syllables (ASW=1.45): 206.835 − (1.015 × 20) − (84.6 × 1.45) = 206.835 − 20.3 − 122.67 = 63.9 (standard difficulty). The challenge is counting syllables accurately — the estimator uses a rule-based syllable counter (vowel groups, prefixes, suffixes) accurate to ±5% for most English prose.