The Gunning Fog Index estimates the years of formal education needed to understand a piece of text on first reading.
Built and maintained by Paul Clark, Redmoon Software · Last verified · Sources · Changes and corrections
Use Gunning Fog when your goal is plain business writing — Robert Gunning created it for newsroom editors who needed a quick way to catch jargon. It is especially useful for evaluating financial reports, legal summaries, and corporate communications where multi-syllable words are easy to slip in unconsciously.
Gunning Fog and SMOG both penalise polysyllabic words, but SMOG uses square-root scaling and is more conservative. Compared to Flesch–Kincaid, Fog is harsher on vocabulary and gentler on sentence length. Use Fog for business; use SMOG for health.
The Gunning Fog index was created by Robert Gunning to identify text that is unnecessarily "foggy" or hard to understand.
It combines average sentence length with the percentage of "complex" words — words of 3+ syllables, not counting words that only reach three syllables through -ed, -es or -ing, or hyphenated compounds.
A score of 12 corresponds to high-school senior level. Time and Newsweek typically score around 11.
A word of 3 or more syllables. This tool skips words that only reach three syllables through a common -ed, -es or -ing ending, and hyphenated compounds. Gunning's original rules also exclude proper nouns and familiar compounds such as "everywhere"; this tool does not, so text full of long names scores slightly higher here.
Aim for under 12 for general audiences. Anything above 16 is considered very difficult to read.
Yes. The Fog index was popularised by Robert Gunning specifically for business and journalistic writing.
Input
The implementation requires consideration of consistency, observability, and serialization.
Output
Gunning Fog Index: 25.8 — Postgraduate.
Nine words in one sentence, five of them 3+ syllables (55.6%). Fog adds the complex-word percentage to the average sentence length and multiplies by 0.4: 0.4 × (9 + 55.6) = 25.8, so dense vocabulary dominates the score.
Fog is 0.4 × (average sentence length + percentage of complex words). Unlike Flesch, the two terms carry equal weight, which makes Fog markedly more sensitive to long sentences and more forgiving of long words than the Flesch family.
A "complex" word here is one of three or more syllables, with exclusions applied — common suffixes that add a syllable without adding difficulty are stripped before the test, so "created" is not treated as complex simply because "-ed" pushed it to three. That exclusion is part of Gunning's original definition, not a shortcut, and without it the index runs high on ordinary past-tense prose.
The output is a US grade level in the same sense as Flesch–Kincaid. Gunning's own guidance was that a Fog above about 12 is heavier than most general readers want, and mainstream news magazines have historically sat near 11.
Fog is the most useful of the classic indices for spotting one specific failure: sentences that keep going. Because sentence length is not down-weighted, a document that scores badly on Fog but acceptably on Flesch is almost always telling you the sentences are too long rather than the words too hard.
It is correspondingly poor on technical writing where the long words are unavoidable and genuinely familiar to the audience. A page about "deoxyribonucleic acid" is not made clearer by finding a shorter word, and Fog will penalise it anyway. Where domain vocabulary dominates, Dale–Chall's familiar-word list is the better instrument, because it asks whether a word is *known*, not whether it is *long*.
The method behind this tool, so you can check it rather than take it on trust.
Gunning, R. (1952). The Technique of Clear Writing. McGraw-Hill.
Introduced the Fog Index, combining sentence length with the proportion of words of three or more syllables.
Classic 0–100 readability score.
Readability scoresU.S. grade level for your text.
Readability scoresHealth & medical text grade level.
Readability scoresNo syllables needed — just characters.