WCAG PDF: Which Success Criteria Bite
On this page
WCAG is written for web content, but it applies to PDFs, and the W3C’s own PDF Techniques explain how to meet each success criterion in a document. Public-sector rules in the US and EU point at WCAG for the documents those bodies publish, so a PDF on a government site is measured against it. This guide lists the criteria that actually bite in a PDF, what each one means in practice, and how to test against them.
Does WCAG apply to PDFs?
It does, and the mechanism is worth understanding. WCAG is a set of success criteria written generically enough to cover “content”, and the W3C maintains a companion document of PDF Techniques that maps each criterion to a PDF-specific way of meeting it. When a rule says an agency must meet WCAG 2.1 AA, that includes the documents the agency publishes, and PDFs are documents.
The version matters and varies by rule. Section 508 incorporates WCAG 2.0 Level AA. The DOJ’s ADA Title II rule sets WCAG 2.1 Level AA, with compliance dates of 26 April 2027 for public entities serving 50,000 or more people and 26 April 2028 for smaller entities and special district governments, as set out on the DOJ web rule. WCAG 2.2 exists and adds criteria on top; meeting 2.1 AA generally satisfies 2.0 AA, because 2.1 is a superset. The criteria themselves are published by the W3C at the WCAG 2.1 specification.
The success criteria that bite in a PDF
Most WCAG criteria are about interaction, which a static PDF does not have. A smaller set does the real work in documents.
| Success criterion | Level | What it means in a PDF |
|---|---|---|
| 1.1.1 Non-text content | A | Meaningful images need alternative text; decorative ones need to be marked as artefacts |
| 1.3.1 Info and relationships | A | Structure and relationships must be programmatically determinable, which means a correct tag tree |
| 1.3.2 Meaningful sequence | A | The reading order must match the logical order of the content |
| 1.4.3 Contrast (minimum) | AA | Text must have enough contrast against its background |
| 2.4.2 Page titled | A | The document needs a title that describes it |
| 3.1.1 Language of page | A | The document language must be set so a reader can pronounce the text |
The first three are where automated tools do the most and still leave work. Info and relationships depends on the tag tree being correct, not merely present. Meaningful sequence depends on the reading order, which is the single most common failure in multi-column layouts. Non-text content depends on alternative text, which is a human judgement every time.
How to test a PDF against WCAG
Testing has two halves, and skipping either one gives a false result.
The automated half is fast. Run the file through a validator such as PAC 2024 or veraPDF, which will report structural failures, missing metadata and untagged content. Both are free and both produce a report worth keeping. The PDF accessibility testing guide covers the tools in more depth.
The manual half is where most failures hide. Read the file with a screen reader and listen to the order. Check that each heading is announced at the right level. Confirm the alt text actually describes the image rather than naming the file. Check contrast on any text over a coloured background. None of this can be automated, because it depends on meaning.
A worked example: a city posts a two-column newsletter as a PDF. A validator passes it, because the file is tagged and the metadata is set. A screen-reader pass reveals the reading order runs across the columns, so the sentence a sighted reader follows down column one is broken by text from column two. The file passes the automated half and fails the manual half, and only the manual test caught it. The fix is in the tag tree, not the content.
WCAG, PDF/UA and the rules that reference them
Three things get mixed together, and keeping them apart makes the work clearer.
WCAG is the success criteria, and it is what the rules name. PDF/UA is ISO 14289, a structural standard for PDFs, and the failure conditions are catalogued in the Matterhorn Protocol published at the PDF/UA Foundation. The two overlap heavily: a PDF/UA-conformant file usually meets the WCAG criteria that apply to documents. Section 508 and ADA Title II are the rules that turn the standards into obligations for particular organisations.
In practice you test once and cite what your rule names. If your obligation is ADA Title II, you cite WCAG 2.1 AA and you can check structure against PDF/UA. If it is Section 508, you cite WCAG 2.0 AA. For the US federal side, see Section 508 PDF, and for the European side, European Accessibility Act PDF.
TagRight fits the first pass, not the last. It tags a PDF in the browser and returns a report of the heading outline it inferred and what it could not fix, which tells you where the manual half of the test needs to focus. It does not write alt text and cannot tag a scan with no text layer, and it is a remediation aid rather than a certification, so verify the result against the criteria above. For a fuller picture of what an accessible document requires, see accessible PDF, and work through the accessible PDF checklist before you publish. You can start with a free document; the plans and pricing cover ongoing work.
Frequently asked questions
Does WCAG apply to PDFs?
Yes. WCAG is written for web content, and the W3C publishes PDF Techniques that explain how to meet each success criterion in a PDF. Public-sector rules in the US and EU point at WCAG for documents, so a published PDF is measured against it.
Which WCAG version applies to PDFs?
It depends on the rule. Section 508 incorporates WCAG 2.0 AA. The DOJ's 2026 ADA Title II rule sets WCAG 2.1 AA. WCAG 2.2 exists and adds criteria, and meeting 2.1 AA generally satisfies 2.0 AA.
Which WCAG criteria are hardest to meet in a PDF?
Info and relationships, meaningful sequence, and non-text content. The first two depend on a correct tag tree and reading order, and the third needs meaningful alternative text, which no automatic tool can write.
Can a PDF meet WCAG AA?
Yes. Conformance is about the content and the structure, and a well-tagged PDF can meet Level AA. The practical test is a validator such as PAC 2024 or veraPDF plus a manual check of alt text and reading order.
Is WCAG the same as PDF/UA?
No. WCAG is a set of success criteria for content accessibility, and PDF/UA is an ISO standard for PDF structure. They overlap heavily, and a PDF/UA-conformant file usually meets the WCAG criteria that apply to documents.