Flesch vs Gunning Fog vs SMOG: which readability formula should you use?

If you measure text readability for a living — editor, accessibility consultant, technical writer, classroom teacher — sooner or later three formulas land on your desk: Flesch Reading Ease, the Gunning Fog Index, and SMOG. They sound interchangeable. They are not.

Each was built for a different audience, with different assumptions about what makes text hard. Picking the wrong one for the job produces scores that look authoritative but mislead. This piece walks through what each formula optimises for, when to choose it, and what to do when they disagree.

The 30-second summary

• Flesch Reading Ease — single 0–100 number, higher is easier. Best when you need an intuitive score to compare two pieces of writing for a general audience.
• Gunning Fog Index — returns a US grade level. Punishes long sentences and 3+ syllable words. Best for business writing, journalism, and finding jargon.
• SMOG — returns a US grade level. Calibrated against 100% comprehension, not approximate. The gold standard for health, medical, and patient-facing writing.

Where each formula came from

Rudolf Flesch published Reading Ease in 1948 as part of his anti-bureaucratic crusade. He wanted a number that anyone could compute by hand and that mapped intuitively to school grades. Hence the 0–100 scale, anchored to 206.835 being the score of pure first-grade vocabulary and 0 being impenetrable academic prose.

Robert Gunning developed the Fog Index in 1952 for newspapers and corporate trainers. His insight: difficulty isn't just about syllables; it's about density of complex words. The Fog Index treats any 3+ syllable word (with some exclusions) as "complex" and combines that ratio with average sentence length.

G. Harry McLaughlin created SMOG (Simple Measure of Gobbledygook) in 1969 specifically because he found Flesch and Fog were too lenient for health materials. SMOG is the only mainstream formula calibrated against 100% comprehension — meaning a SMOG-7 score is what a typical 7th-grader will fully understand, not approximately get the gist of.

When to choose Flesch Reading Ease

Use Flesch when your reader is a general adult and you need a single, recognisable number. Microsoft Word ships it natively. Marketing teams understand "score of 70+". Editors comparing two drafts can spot the difference at a glance.

Flesch is also the right choice when communicating about readability to non-experts. "Our copy scores 65" is meaningful to people who have never heard of Gunning Fog. The formula's simplicity is a feature: words and syllables, nothing else.

Where Flesch falls short: technical content with unavoidable long words ("photosynthesis", "implementation") will score low even with short, clear sentences. The number can also be gamed — break every long sentence in half and your Flesch rises without your meaning improving.

When to choose Gunning Fog

Use Gunning Fog for business writing, journalism, legal summaries, and financial reports. The Fog Index was built to catch the kind of unconscious jargon that corporate writing accumulates — "facilitate", "stakeholder", "operationalise". It penalises these specifically.

If you're an editor working with a writer who keeps slipping into bureaucratic register, Fog is your friend. It exposes complex vocabulary in a way that Flesch (which rewards short syllables generically) does not.

Where Fog falls short: it treats every 3+ syllable word as complex, including perfectly familiar ones like "actually" or "interesting". And it's English-only — the syllable rules don't generalise.

When to choose SMOG

Use SMOG for any health, medical, safety, or regulatory writing. The National Institutes of Health, the CDC, and most patient-education guidelines recommend SMOG specifically. The reason: SMOG is the only formula calibrated against full comprehension, not approximate understanding.

When someone needs to follow medication instructions exactly, "they probably got the gist" is not good enough. SMOG-7 means a typical 7th-grader will actually understand it. That extra rigour translates to scores about 1–2 grades higher than Flesch–Kincaid on the same text.

SMOG is also more conservative because it uses a square-root scaling on polysyllable counts. A few extra long words have a noticeable but bounded effect — it can't spike from a single outlier sentence the way Fog can.

What to do when they disagree

It's common to run all three and get genuinely different answers. Here's what disagreement usually means:

• Flesch high, Fog high: text has unavoidable jargon but short sentences. The vocabulary issue is real; consider a glossary or rewrites for unfamiliar terms.
• Flesch low, Fog low: long sentences but ordinary words. Break sentences; the vocabulary is fine.
• SMOG much higher than Flesch–Kincaid: many 3+ syllable words. For health/safety writing, trust SMOG and simplify.
• Fog much higher than Flesch–Kincaid: a few sentences are pulling up the complex-word ratio. Find and edit them rather than rewriting the whole text.

What none of them measure

All three formulas measure mechanical readability: how the words and sentences look on the page. They do not measure:

• Vocabulary familiarity — "epistemology" and "thinking" count the same once "thinking" is past two syllables. For familiarity, use Dale–Chall.
• Logical flow — sentences can be short and clear yet hang together incoherently.
• Quality — Hemingway scores well on Flesch; so does a poorly written email.
• Cultural register — formal vs. casual is invisible to syllable counters.

Treat readability formulas as smoke detectors, not judges. They flag potential problems for human review — they don't decide if writing is good.

A practical workflow

1. Run Flesch Reading Ease first for a quick sanity check and a number you can share.
2. If the text is health, medical, or safety-related, run SMOG and trust it over Flesch.
3. If the text is business or corporate, run Gunning Fog to spot jargon density.
4. If the scores disagree by more than two grades, read the highest-scoring formula's worst-offending sentences and edit them specifically.
5. Re-run, repeat. Stop when you've matched your target audience, not when you've hit a particular number.

The formulas are tools, not arbiters. Used well, they catch the lazy long sentences and bureaucratic word choices that slip past human review. Used badly, they generate a false sense of precision that papers over real writing problems. The discipline is in knowing which one to pick — and when to stop trusting it.