Guide
Acrobat accessibility checker: every rule explained
What each of the 32 rules in Acrobat's Full Check means, which PDF/UA condition it matches, and how to fix it — from Tab order failed to Logical Reading Order.
About 16 minutes · moderate
How to read the report
Acrobat Pro’s Accessibility → Accessibility Check → Full Check runs 32 rules in seven groups and reports each as Passed, Failed, Needs manual check or Skipped. Right-click any rule for Fix, Explain or Check again.
Two things are worth knowing before you start:
- “Passed” everywhere is not “accessible”. Some rules can only ever say Needs manual check, and Acrobat does not test many of the Matterhorn Protocol’s 136 failure conditions. It also does not test whether Hebrew or Arabic text is stored backwards — see that failure.
- The automatic Fix is a starting point. It is fine for the title or the language. For reading order, tables and alternate text it produces something that passes the rule without being right.
Below, each rule: what it means, the PDF/UA condition it corresponds to where there is one, and the fix.
Document
Accessibility permission flag. The file’s security settings stop assistive technology from reading the text. Corresponds to 26-002. Fix: File → Properties → Security, and allow content copying for accessibility, or remove the security.
Image-only PDF. There is no text in the file, only pictures of pages — a scan. Fix: OCR it, then check the result by reading it. Remediating a scanned PDF →
Tagged PDF. The file has no structure tree. Nothing else can be fixed until it has one. Fix: re-export from the source with tagging on, or Add Tags to Document as a first pass. Tagging an untagged PDF →
Logical Reading Order — Needs manual check. Always manual. Software cannot know whether the order is the order you meant. Corresponds to 09-001. Fix: listen to it, or read the tags top to bottom. Fixing reading order →
Primary language. The document declares no language, so a screen reader guesses the voice. Corresponds to 11-001. Fix: File → Properties → Advanced → Language.
Title. No document title, or the window shows the file name instead. Corresponds to 06-003 and 07-001. Fix: File → Properties → Description → Title, then Initial View → Show: Document Title. Setting the title →
Bookmarks. A long document has none. PDF/UA sets no machine test here (27-001); WCAG 2.4.5 is where it bites. Fix: bookmarks from the headings. Headings and bookmarks →
Color contrast — Needs manual check. Always manual, and a WCAG requirement (1.4.3) rather than a PDF/UA one. Colour contrast in PDFs →
Page content
Tagged content. Some text, images or paths are neither tagged nor marked as artifacts, so a screen reader either skips them or reads them at an unpredictable point. Corresponds to 01-005. Fix: tag real content, mark decoration as artifacts.
Tagged annotations. A link, comment or form field is not in the structure tree. Corresponds to 28-002 and 28-011. Fix: tag each annotation in its element — a link inside <Link>.
Tab order. A page with annotations does not say that tab order follows the structure. Corresponds to 28-008. Fix: Page Thumbnails → select all → Page Properties → Tab Order → Use Document Structure.
Character encoding. Some characters cannot be mapped to Unicode, so a screen reader cannot say them. Corresponds to 10-001. Fix: re-export with the fonts embedded; a missing ToUnicode map is rarely repairable by hand.
Tagged multimedia. Audio or video is not tagged. Rare in documents.
Screen flicker, Scripts, Timed responses, Navigation links. WCAG concerns — flashing content, scripts that stop keyboard access, time limits, repeated links. Uncommon in ordinary documents; check them manually if your file has scripts or media.
Forms
Tagged form fields. A form field is not in the structure tree. Corresponds to 28-010. Fix: tag each field in a <Form> element.
Field descriptions. A field has no tooltip, so it is announced as “edit text” with no label. Corresponds to 28-005. Fix: the field’s Properties → General → Tooltip, written as the label a sighted user reads. Accessible PDF forms →
Alternate text
Figures alternate text. A <Figure> has no description. Corresponds to 13-004. Fix: Accessibility → Set Alternate Text, or mark it decorative. Alt text that works →
Nested alternate text. An element has alternate text and so does something inside it; the inner one will never be read. Fix: keep the description on one level.
Associated with content. Alternate text sits on an element that holds no content, so nothing ever triggers it. Fix: move it to the element that contains the image.
Hides annotation. Alternate text on an element that contains a link or field hides that annotation from the reader. Fix: take the alternate text off the parent.
Other elements alternate text. Something other than a figure — often a formula — needs a description and has none.
Tables
Rows and TH and TD. The table’s grammar is broken: rows outside a table body, cells outside rows. Corresponds to 09-004. Fix it before anything else; Acrobat’s Table Editor will not open a broken table.
Headers. A table has no header cells, or they carry no Scope. Corresponds to 15-003. Tagging tables →
Regularity. Rows have different numbers of cells. Usually merged cells without row or column spans.
Summary. No table summary. PDF/UA-1 does not require one; Acrobat asks. Many teams add a short one or accept the flag.
Lists
List items and Lbl and LBody. The list’s structure is broken — an item outside a list, a label outside an item. Corresponds to 09-005. Fix: <L> → <LI> → <Lbl> and <LBody>.
Headings
Appropriate nesting. Heading levels skip on the way down — H1 straight to H3. Corresponds to 14-003. Skipped heading levels →
What Acrobat does not check
It does not test most of PDF/UA’s font and metadata conditions, it cannot decide any condition that needs judgement, and it does not look at text direction. Run veraPDF or PAC as well, and for Hebrew or Arabic documents, check direction separately — Taggart’s free checker does, with no account.
Frequently asked questions
Acrobat says my PDF passed. Is it accessible?
It means none of the 32 rules Acrobat checks failed automatically. Several of them always need a person — Logical Reading Order and Color contrast report 'Needs manual check' on every file — and Acrobat does not test many of the Matterhorn Protocol's 136 conditions at all. A pass is where checking starts, not where it ends.
What does 'Needs manual check' mean?
That the rule cannot be decided by software. Acrobat is telling you, correctly, that it does not know whether the reading order is right or the contrast is sufficient. The status stays until you decide and mark it Passed or Failed yourself.
Is Acrobat's checker the same as PDF/UA validation?
No. It overlaps with PDF/UA but is its own list, and it tests things PDF/UA does not require (a table Summary) while missing things PDF/UA does (most of the font and metadata conditions). For a PDF/UA verdict, use veraPDF or PAC as well.
Does Acrobat check Hebrew and Arabic text direction?
No, and neither does PDF/UA. A Hebrew PDF stored in visual order — reversed — can pass every one of Acrobat's rules and still read backwards to a screen reader. That needs a separate check.
Related guides
- How to check if a PDF is accessible: four free checkersCheck PDF accessibility free, online or on your own machine: four checkers against PDF/UA and WCAG, what each catches, and the failures none of them see.
- How to fix PDF reading orderA working method for finding and repairing reading order, free in Acrobat and faster in Taggart, including the two-column case that costs the most time.
- How to tag a table in a PDF for accessibilityTable structure, header scope, and the point where Scope stops working and Headers references become necessary.
- How to add and write alt text for images in a PDFAdding alt text to a PDF image starts with deciding whether it is informative. Then what to say, what to leave out, how to add it free, and where AI helps.
Related failure conditions
- 28-008 A page containing an annotation does not contain a Tabs entryAny page carrying annotations must declare a tab order. PDF/UA-1 requires /Tabs /S — tab order follows the structure tree.
- 10-001 Character code cannot be mapped to UnicodeA glyph is painted whose character code has no Unicode equivalent reachable through the font's ToUnicode CMap, encoding, or glyph names. The text renders correctly but cannot be extracted.
- 11-001 Natural language for text in page content cannot be determinedNo /Lang is in effect for some text: neither on the catalog, nor on an enclosing structure element, nor on a marked-content sequence.
- 13-004 Figure tag alternative or replacement text missingA Figure element has neither /Alt nor /ActualText. Every Figure must have one or the other; if the image is decorative it should be an artifact instead of a Figure.
- 09-004 A table-related structure element does not conform to ISO 32000-1, Table 337The table grammar is broken: a TD outside a TR, a TR outside Table/THead/TBody/TFoot, a Table with no rows, or rows that span different numbers of columns.
- 15-003 A TH cell does not contain a Scope attributeHeader cells must declare whether they head a row, a column, or both, unless the table instead uses explicit /Headers ID references on every data cell.
- 09-005 A list-related structure element does not conform to ISO 32000-1, Table 336The list grammar is broken: an LI outside an L, an Lbl or LBody outside an LI, a list item with no LBody.
- 14-003 Numbered heading levels in descending sequence are skippedThe heading hierarchy jumps a level going down — H1 straight to H3. Going back up any number of levels is fine; going down must be one level at a time.
- 26-002 The 10th bit position of the P entry is falseBit 10 of the permissions flags is the "extract for accessibility" permission. Clearing it tells assistive technology it may not read the file.
- 01-005 Content is neither marked as Artifact nor tagged as real contentEvery text-showing, image-painting and path-painting operator on a page must sit either inside a marked-content sequence that maps to the structure tree, or inside an artifact sequence. Content that is in neither is orphaned.
Check your own file. Taggart's validator is free and unlimited — no page cap, no watermark, and no account at all up to 60 pages.
Last updated .