Readability and Accessibility: WCAG Compliance, Screen Readers, and Readability Formulas

Readability formulas measure surface-level text characteristics—sentence length, syllable count, word frequency—but readability and accessibility WCAG conformance covers much more: structure in code, text alternatives, contrast, and keyboard and screen-reader access. Passing both requires understanding where formulas succeed and where they fail.

Introduction

Many writers and compliance teams treat readability scores and WCAG accessibility as interchangeable targets. They are not. A document can score well on the Flesch-Kincaid Grade Level scale and still fail WCAG 2.1 AA because it lacks proper heading hierarchy, has poor color contrast, or includes charts with no text alternative.

Conversely, a document can meet WCAG structural requirements and still be difficult to parse for readers with cognitive disabilities, dyslexia, or low vision if the underlying prose is convoluted.

This guide bridges that gap. It explains what WCAG actually mandates about readability, how screen readers interact with text formatting and punctuation, and which readability formulas—like Flesch-Kincaid and the Gunning Fog Index—correlate with real accessibility outcomes. You'll learn concrete compliance targets for each WCAG level, see why cognitive accessibility extends beyond grade-level scores, and walk through a healthcare example that passes both formula benchmarks and accessibility audits.

Whether you're writing patient information, compliance documentation, or public-facing content, understanding the relationship between readability metrics and WCAG requirements ensures your content works for humans and machines.

The Gap Between Readability Formulas and True Accessibility

Readability formulas predict how easily people can understand prose. Rudolf Flesch published the Reading Ease formula in 1948. In 1975, J. Peter Kincaid and colleagues turned it into the Flesch-Kincaid Grade Level for the U.S. Navy. Gunning Fog Index, created by Robert Gunning in 1952, focuses on complex words. These metrics work by counting syllables and sentence length—mechanical operations that take seconds to compute.

WCAG (Web Content Accessibility Guidelines), published by the W3C and updated most recently in 2023–2024 with WCAG 3.0 drafts in progress, addresses a broader mandate: ensuring content is perceivable, operable, understandable, and robust for all users, including those with disabilities. WCAG does reference readability, but it does so alongside requirements for color contrast, keyboard navigation, alt text, form labeling, and heading structure.

The critical difference: a readability formula cannot detect that a long sentence lacks proper punctuation breaks for screen readers, that a numbered list lacks semantic markup, that a table has no header row, or that colored text conveys meaning without a text alternative.

Additionally, why grade level isn't reading level explores why a 6th-grade Flesch-Kincaid score does not guarantee a 6th-grader can understand the content. Domain vocabulary, prior knowledge, conceptual density, and layout all matter. WCAG addresses many of these factors; readability formulas largely ignore them.

WCAG 2.1 and 3.0 Readability Requirements: What the Standards Actually Say

WCAG 2.2 is the current W3C Recommendation. WCAG 2.1 and 2.0 are still the versions named in some laws. None of them mandates a readability score. WCAG has three conformance levels: A, AA and AAA.

WCAG 2.1 Level A includes no readability criterion. Its requirements cover text alternatives, keyboard access, and not relying on color alone to convey meaning.

WCAG 2.1 Level AA does not include a reading-level requirement either. Level AA is the level most laws and procurement rules point to, including Section 508, which adopts WCAG 2.0 Level A and AA.

WCAG 2.1 Level AAA is where reading level appears, as Success Criterion 3.1.5, "Reading Level." It states: "When text requires reading ability more advanced than the lower secondary education level after removal of proper names and titles, supplemental content, or a version that does not require reading ability more advanced than the lower secondary education level, is available."

Translation: WCAG defines the lower secondary level as about nine years of school, roughly the end of US grade 9. If your prose needs more reading ability than that, Level AAA asks for a simplified version or supplemental explanation. This is not about banning complex writing—it is about ensuring users with cognitive disabilities or language barriers can access equivalent information.

Level AAA also holds WCAG's other language-clarity criteria: 3.1.3 Unusual Words, 3.1.4 Abbreviations, and 3.1.6 Pronunciation. W3C does not recommend requiring AAA conformance for entire sites, because some content cannot meet every AAA criterion.

WCAG 3.0 is still a working draft. It is organised around outcomes, and it is not yet a standard you can conform to.

Which version applies depends on who you are. US federal agencies follow WCAG 2.0 Level A and AA under Section 508. The Justice Department's 2024 rule for state and local governments adopts WCAG 2.1 Level AA. For private businesses, the ADA names no technical standard. None of these sets a reading level. Writing at or below the lower secondary level is still good practice, and it is what Level AAA asks for: provide plain-language summaries for specialized content, and ensure proper semantic structure.

How Screen Readers Interact With Sentence Length and Punctuation

Screen readers convert text to synthesized speech. Long sentences without punctuation breaks create distinct accessibility problems:

Run-on sentences and ambiguous boundaries. A screen-reader user hears: "The patient should take the medication with food which can reduce nausea and also improves absorption so do not take on an empty stomach." Where does one clause end? Where does advice change? A human reader's eye can scan and parse; a screen-reader user can step back by word or line, but it takes more effort.

Semantic punctuation. Periods, commas, colons, and hyphens are not merely visual—they signal grammatical boundaries. Many screen readers pause at punctuation. Removing or misusing punctuation degrades comprehension for both sighted and screen-reader users.

Conjunction overload. Sentences strung together with "and," "but," "however," and "therefore" without structural markers (headings, lists, line breaks) force screen-reader users to hold multiple concepts in working memory. A sentence like "Patients with hypertension and diabetes and a family history of stroke should monitor blood pressure weekly and report readings above 140/90 and seek immediate care if experiencing chest pain or dizziness" is hard to parse by ear.

Optimal structure for screen readers: Keep sentences under 20 words on average. Use lists instead of prose for multiple conditions or steps. Use active voice. Break complex thoughts into separate sentences. Ensure each sentence stands alone or connects clearly to the prior one.

Readability formulas explained can detect long sentences, but they cannot evaluate whether punctuation and conjunctions are placed optimally for audio consumption.

Grade-Level Targets for WCAG's Reading-Level Criterion

WCAG names no readability formula. To stay safely under the lower secondary threshold that Success Criterion 3.1.5 (Level AAA) uses, these are reasonable working targets. They are this site's guidance, not numbers from WCAG:

• Flesch-Kincaid Grade Level: 6–8
• Gunning Fog Index: 8–10
• Flesch Reading Ease: 60–70 (higher is easier)
• SMOG Index: 6–8

These are target ranges, not hard cutoffs. A Flesch-Kincaid score of 7.2 is comfortably under the threshold, and 8.5 is close to it. At 10 or above, a site aiming for Level AAA would need a simpler version or supplemental plain-language content.

In practice:

• Healthcare content (patient information, discharge instructions): aim for Flesch-Kincaid 5–7. Patients under stress retain less; lower is better.
• Government forms and public notices: aim for Flesch-Kincaid 6–8, or lower for benefits notices read under stress. See plain language compliance standards for what the law actually requires.
• Legal and compliance documents: Flesch-Kincaid 8 is a realistic target, but larger font, chunking, and glossaries become more important.

If you pursue WCAG Level AAA, Success Criterion 3.1.5 applies: any text that needs more than lower-secondary reading ability must come with a simpler version or supplemental content. Writing to Flesch-Kincaid 5–6 in the first place avoids that extra work, and it helps readers with cognitive disabilities or dyslexia.

Why multiple formulas? Because they measure different aspects. Flesch-Kincaid emphasizes syllables and sentence length. Gunning Fog counts words of three or more syllables. SMOG also counts those words and is set to full comprehension, so it usually runs higher than Flesch-Kincaid. Use all three; if they agree, your text is predictably readable. If they diverge, investigate: complex words? Long sentences? Unfamiliar domain terms?

Cognitive Accessibility: Beyond Grade-Level Scores

Grade-level readability scores cannot measure cognitive load—the mental effort required to parse, hold, and act on information. Two pieces of content can both score Flesch-Kincaid 7.0 and present vastly different cognitive burdens.

Consider:

Version A (grade 25.0): "Patients with a history of myocardial infarction, hypertension, hypercholesterolemia, or diabetes mellitus should be screened for atrial fibrillation using 24-hour Holter monitoring or event monitoring based on symptom frequency."

Version B (grade 7.6): "Your doctor will check your heart rhythm if you have had a heart attack, high blood pressure, high cholesterol, or diabetes. We use a 24-hour heart monitor or a portable device you carry. Your doctor picks the right test based on how often you feel symptoms."

Version B scores far lower, and it is easier in ways the score does not show: it uses familiar words ("doctor," "heart attack" rather than "myocardial infarction"), shorter sentences, and logical chunking. Version B also uses the second person ("your"), which speaks to the reader directly.

Cognitive accessibility strategies that readability formulas miss:

• Consistent terminology. Use "blood pressure" throughout; do not alternate with "BP" or "hypertension measurement." Synonyms force working memory to map equivalences.
• One idea per sentence. Even if a sentence scores low on Flesch-Kincaid, multiple clauses buried in it overload cognitive load.
• Visual hierarchy and whitespace. Lists, subheadings, and margins are not readability metrics, but they make text easier to scan. WCAG addresses structure; readability formulas do not.
• Concrete language. "Take your medicine every morning after breakfast" is easier to execute than "Administer medication following a morning meal." Both may score similarly on grade level, but one is actionable.

For patient information readability, cognitive accessibility means testing with actual users, not just scoring. A document that passes Flesch-Kincaid 5.0 and WCAG AA but confuses users with dense medical terminology is not truly accessible.

Real Example: Healthcare Form Optimized for WCAG and Readability Scores

Original version (non-compliant):

"All patients presenting with complaints related to cardiovascular dysfunction, including but not limited to chest discomfort, dyspnea, arrhythmias, or syncope, must complete a comprehensive health history questionnaire prior to examination and provide documentation of prior diagnostic studies, pharmaceutical interventions, and familial disease states."

Metrics: one 43-word sentence. Flesch-Kincaid 28.9. Flesch Reading Ease 0, the bottom of the scale.

Revised version (plain language, with list structure):

"Please tell us about your heart health. Have you had any of these?

• Chest pain or pressure
• Shortness of breath
• Heart racing or skipping
• Fainting or dizziness

Bring any test results or medication lists from previous visits."

Metrics: Flesch-Kincaid 5.2 on this site's calculator, well under the lower secondary level that WCAG 3.1.5 (Level AAA) refers to.

What changed:

1. Replaced medical jargon ("cardiovascular dysfunction," "dyspnea," "syncope") with plain language ("heart racing," "shortness of breath," "fainting").
2. Broke a 43-word sentence into a short opener, a list, and a 10-word instruction.
3. Used second person ("Have you had") instead of passive ("patients presenting").
4. Removed "prior diagnostic studies" and "pharmaceutical interventions"—replaced with "test results" and "medication lists."

The revised version also works better with WCAG structure: the list is marked up as a real list, and the instruction is short and clear. A screen-reader user hears one item at a time.

This revision is plain-language text with proper list structure. WCAG Level AA has no reading-level test, so readability is not something a page passes at AA. Accessibility testing tools check other things, such as alt text, form labels, color contrast, and keyboard navigation.

Heading Structure, Chunking, and Why Formula Metrics Miss Half the Picture

Readability formulas analyze running text. They ignore headers, lists, subheadings, and white space—all critical to accessibility and actual readability.

A document with:

• Poor heading hierarchy (missing <h2>, jumping from <h1> to <h4>)
• Long paragraphs (400+ words)
• No bulleted or numbered lists
• No visual breaks

…will score well on Flesch-Kincaid if individual sentences are short, but it will fail accessibility audits and frustrate users because the structure is inaccessible.

WCAG 1.3.1 (Level A) requires that structure be available in code. In practice:

• Headings marked up as headings. Keeping levels in order is good practice.
• Semantic markup (<h1>, <h2>, not styled <p> tags).
• Lists marked as <ul> or <ol>, not prose with commas.
• Form fields with a programmatic label, such as a <label> tag.
• Tables with header rows (<th> tags).

A screen reader depends on this structure to orient users and allow navigation. "Jump to next heading" is a standard screen-reader command; if headings are not marked, the command fails.

Best practice: Use WCAG's heading and list guidelines to break text into digestible chunks. Then measure readability of each chunk. A section with a subheading, a 2-3 sentence intro, a bulleted list, and a concluding sentence is far more accessible than the same information in a 6-sentence paragraph—even if both score Flesch-Kincaid 6.0.

Testing Tools: Which Readability Formulas Correlate With Accessibility Compliance

Not all readability tools are built equally. Some integrate accessibility checks; most do not.

Readability-only tools (detect formula scores but not WCAG compliance):

• Hemingway Editor: Highlights long sentences, passive voice, and complex words. Good for drafting. Does not check heading structure or semantic markup.
• Flesch-Kincaid (Microsoft Word, Google Docs plugins): Fast, familiar. Limited to grade-level estimates. No color contrast or alt-text checking.
• Grammarly: Offers readability and tone suggestions.

Accessibility + readability integrated tools:

• WAVE (WebAIM): Checks WCAG contrast, alt text, heading hierarchy, and form labels. It does not score reading level.
• Axe DevTools (Deque): Automated WCAG rule checks. It does not score reading level.

Most relevant for balanced compliance: Use a combination approach. Run your text through Hemingway Editor or Microsoft Word for grade-level scores and sentence-length detection. Then run the final HTML/web content through WAVE or Axe DevTools to ensure heading structure, contrast, alt text, and form labeling are WCAG-compliant.

For plain-language writers focused on government and public communication, the plain language compliance standards page outlines both readability benchmarks and regulatory accessibility requirements.

Bottom Line

Readability formulas and WCAG compliance are complementary but distinct. Readability formulas predict how easily humans parse prose and are the practical way to check WCAG's one reading-level criterion, 3.1.5, which sits at Level AAA and draws its line at the lower secondary level. WCAG also requires semantic structure, text alternatives, and keyboard access—things formulas cannot measure. To optimize for both, target a Flesch-Kincaid score of 5–8 depending on your audience, ensure proper heading hierarchy and list markup, test with accessibility tools like WAVE or Axe DevTools, and—for healthcare, legal, and other high-stakes domains—validate your work with real users, including those with dyslexia, low vision, and cognitive disabilities. The goal is not a perfect formula score; it is content that works for everyone.