Flesch-Kincaid vs Gunning Fog vs SMOG: Which Readability Formula Should You Use?

Every readability formula is, at its core, a shortcut. Researchers in the mid-twentieth century needed a fast way to estimate how hard a piece of writing was to understand — before computers, before eye-tracking studies, before neuroscience could weigh in. So they counted syllables. They counted sentences. They ran regressions against reading tests and called it a day.

Decades later, those shortcuts still show up in content dashboards, editorial style guides, and SEO tools everywhere. Flesch-Kincaid. Gunning Fog. SMOG. If you've spent any time thinking about content quality, you've probably run your writing through at least one of them and wondered: how much should I actually trust this number?

The honest answer is: it depends on which formula you're using, and what you're trying to measure. Let's go through all three.

Flesch-Kincaid: The Default, For Better or Worse

Rudolf Flesch developed the original Reading Ease formula in 1948. J. Peter Kincaid later adapted it for the U.S. Navy in the 1970s to produce the Grade Level variant most people use today. The math behind it looks like this:

Flesch-Kincaid Grade Level = (0.39 × words/sentences) + (11.8 × syllables/words) − 15.59

Two inputs: average sentence length and average syllables per word. That's it. The output maps to a U.S. grade level — a score of 8 means an eighth-grader should be able to read it comfortably, a score of 12 means high school senior, and so on.

The reason this formula dominates is simple: it's baked into Microsoft Word, Google Docs extensions, Yoast SEO, Hemingway App, and dozens of content platforms. It's the path of least resistance. But that ubiquity obscures some real limitations.

Flesch-Kincaid is extremely sensitive to sentence length. A single 45-word sentence will spike your score dramatically even if every word in it is "the," "and," and "cat." Conversely, you can write genuinely dense, jargon-heavy content in short punchy sentences and end up with a Grade 6 score that makes your article look approachable when it absolutely is not. The formula doesn't care what your words mean. It only cares how many syllables they contain.

What it rewards: Short sentences. Short words. Minimalist punctuation (fewer commas = fewer clause breaks = shorter apparent sentence structure).

What it penalizes: Technical terminology, even when the audience expects and understands it. Complex proper nouns. Any word with three or more syllables, regardless of whether readers actually know it.

Best use case: Consumer-facing web content, marketing copy, general health information, news articles aimed at a broad audience. If you're writing for someone who might be skimming on a phone between meetings, Flesch-Kincaid is a reasonable guardrail.

Gunning Fog: Harder to Game, Still Imperfect

Robert Gunning introduced his Fog Index in 1952, partly because he thought existing formulas were too easy to manipulate. His formula:

Fog Index = 0.4 × ((words/sentences) + 100 × (complex words/words))

The key difference is that "complex words" here means words with three or more syllables — but with some important exclusions. Proper nouns don't count. Compound words made from simple words ("nevertheless," "however") are excluded in strict applications. Verb forms that become three syllables due to -ed or -es suffixes are sometimes excluded too.

The output is also a grade level, and the general guidance is that anything above Fog 12 is "foggy" — too dense for most general-audience writing. Newspapers tend to aim for Fog 10–12. Academic journals often hover around Fog 15–20.

Gunning Fog is harder to game than Flesch-Kincaid because it specifically targets polysyllabic vocabulary, which tends to correlate more directly with genuine comprehension difficulty. If your writing is full of words like "methodology," "implementation," "differentiation," and "accountability," Fog will catch that in a way that plain syllable-averaging sometimes doesn't.

The limitation is the exclusion rules. In practice, most automated tools don't apply the proper-noun or compound-word exceptions consistently. Run the same paragraph through five different Fog calculators and you may get five slightly different scores. The formula is theoretically sound but operationally messy.

What it rewards: Precise, concrete vocabulary. Strong verbs over nominalizations. Shorter words where shorter words genuinely work.

What it penalizes: Technical jargon, even discipline-specific terms the audience knows perfectly well. Long proper nouns (a piece about "Worcestershire" or "Czechoslovakia" will suffer).

Best use case: Business writing, corporate communications, B2B content where you want to signal clarity without dumbing things down. If you're writing an annual report, a white paper, or a product explainer, Gunning Fog is probably more diagnostic than Flesch-Kincaid.

SMOG: The Formula Epidemiologists Actually Use

G. Harry McLaughlin developed SMOG (Simple Measure of Gobbledygook — yes, really) in 1969. It takes a different philosophical approach: instead of averaging across an entire document, it samples.

SMOG Grade = 3 + √(polysyllable count in 30 sentences)

Traditionally, you take 10 sentences from the beginning, 10 from the middle, and 10 from the end of a document. You count every word with three or more syllables across those 30 sentences. Take the square root. Add 3. Done.

SMOG doesn't care about sentence length at all. It is purely a vocabulary measure. And this turns out to matter a lot, because research has consistently shown that word difficulty is a stronger predictor of reading comprehension than sentence length — particularly for readers with limited literacy or those reading in a second language.

This is why SMOG has become the standard in public health writing, patient education materials, government documents, and anything that genuinely needs to reach a wide reading-ability range. The National Institutes of Health recommends SMOG. The Centers for Disease Control uses it. If you're writing a medication guide or a voter information pamphlet, this is the formula that health literacy researchers will judge you by.

The catch: SMOG requires at least 30 sentences to function correctly. It performs poorly on short content — a 200-word landing page, a product description, an FAQ answer. It was designed for pamphlets, reports, and longer documents, and applying it to short-form content produces unstable results.

What it rewards: Plain vocabulary. Using "use" instead of "utilize," "start" instead of "initiate," "help" instead of "facilitate." Word choices that favor clarity over precision signaling.

What it penalizes: Any technical terminology, even unavoidable medical or legal terms. Content that necessarily requires field-specific language will always look "foggy" under SMOG, regardless of how clearly it's actually explained.

Best use case: Health education, government communications, legal plain-language initiatives, accessibility compliance, anything targeting readers who may have lower literacy levels or who are reading in their non-native language.

Putting Them Side by Side

The formulas agree more often than they disagree — if your writing is genuinely dense, all three will flag it. But they diverge in interesting ways. A document with long but simple sentences (think: conversational storytelling) will score worse on Flesch-Kincaid and Gunning Fog than on SMOG. A document with short sentences but heavy technical vocabulary will score worse on SMOG than on Flesch-Kincaid. Knowing which direction a formula tilts helps you understand what it's actually telling you.

The Real Limitation None of Them Solve

All three formulas share the same foundational blind spot: they measure surface features, not comprehension. They cannot tell whether your logical structure makes sense, whether you've defined your terms, whether your examples actually illuminate your point, or whether you've assumed knowledge your reader doesn't have.

A paragraph of perfectly simple sentences that presents ideas in incoherent order will score Grade 5 on Flesch-Kincaid while confusing every reader who encounters it. Meanwhile, a paragraph explaining a genuinely complex concept with precise vocabulary, clear analogies, and solid logical flow might score Grade 14 while being perfectly understandable to its intended audience.

This is the core problem with using readability scores as editorial gatekeepers rather than as one diagnostic signal among many. They're most useful as a check against your own habits — if you have a tendency to write sprawling 60-word sentences, Flesch-Kincaid will catch it. If you're prone to nominalization soup ("the implementation of the utilization of"), Gunning Fog will flag it. If you're writing for a mixed-literacy public audience, SMOG will tell you where your vocabulary is outpacing your readers.

Use them as you'd use spell-check: a fast pass that catches the obvious stuff, not a replacement for thinking clearly about who you're writing for and what they need to understand.

Which One Should You Actually Choose?

Use Flesch-Kincaid when you're writing for the web and want a quick pass that integrates with every tool you already have. It's imperfect but consistent and widely benchmarked.

Use Gunning Fog when you're writing professional, B2B, or business content and want a formula that's slightly more sensitive to vocabulary complexity than sentence structure. It's better suited to longer-form pieces where jargon is the real risk.

Use SMOG when your audience includes people who may struggle with reading, when compliance with health literacy or plain-language standards matters, or when you're writing anything that must genuinely reach the widest possible population. It's not a general-purpose tool, but for the use cases it was designed for, it's the most accurate of the three.

And when in doubt: read your draft out loud. If you stumble, so will your reader. No formula catches what a single careful read-through will.