Includoc, home

What is a tagged PDF? Tags, the tag tree and how to check them

A tagged PDF has hidden structure that tells screen readers what each part is. Learn the common tags, artifacts, role maps and how to view and fix tags.

Last updated

A PDF page lifted to reveal the tag tree structure underneath

A tagged PDF is a PDF that contains a hidden layer of structure, called tags, that labels each piece of content: this is a heading, this is a paragraph, this is a list, this is a table, this image shows a detour map. Screen readers, braille displays, read-aloud tools and reflow use the tags to present the document in a logical order. Tags are required for an accessible PDF, but a PDF isn't accessible just because it's tagged. The tags also have to be correct.

What tags do

A PDF page is a set of drawing instructions: put these letters here, draw a line there, place this image in the corner. That's all a PDF needs to look right on screen or paper, but it says nothing about meaning. Without tags, a screen reader has to guess which text is a heading, where a column ends and whether a run of numbers is a table.

A tagged PDF adds a separate structure, the tag tree (also called the structure tree). It groups the page content into elements such as headings, paragraphs, lists and tables, and puts them in reading order. Each tag points to the content it describes. Tags never change how the document looks.

The technical detail, briefly. The tag tree starts at the document's structure tree root. Content on each page is linked to its tag with a marked-content identifier, and the document's catalog declares that the file is tagged. The PDF specification (ISO 32000) defines the standard tag types, and PDF/UA (ISO 14289) sets out how they must be used for accessibility. You don't need to know this to fix tags, but it explains why tags live in a separate panel from the page content.

What a tag tree looks like

Here's a simplified tag tree for the first page of a meeting agenda:

Document
  H1      "City Council Regular Meeting, March 9, 2027"
  P       "6:00 p.m., Council Chambers, City Hall"
  H2      "1. Call to order"
  H2      "2. Consent calendar"
  L
    LI
      Lbl    "a."
      LBody  "Approve minutes of February 23, 2027"
    LI
      Lbl    "b."
      LBody  "Accept the quarterly investment report"
  H2      "3. Public hearing: Oak Avenue bike lanes"
  Figure  Alt: "Map of the proposed bike lanes on Oak Avenue"
  Link    "Staff report: Oak Avenue bike lanes (PDF)"

A screen reader follows this tree from top to bottom. Because the headings are tagged, someone can jump straight to item 3. Because the list is tagged, they hear "list, 2 items". Because the map has alt text, they know what it shows.

Common PDF accessibility tags

TagWhat it meansNotes
DocumentThe root that holds everything elseMost tag trees start here
Part, Sect, DivGroups of related contentFor organization; most screen readers don't announce them
H1 to H6Headings, levels 1 to 6Go down one level at a time, and don't mix them with the generic H tag
PParagraphThe most common tag
L, LI, Lbl, LBodyList, list item, label (the bullet or number) and list bodyA list holds items; each item holds a label and a body
Table, TR, TH, TDTable, table row, header cell, data cellHeader cells need a scope; THead, TBody and TFoot can group rows
FigureAn image, chart or graphicNeeds alt text, or should be an artifact if it's decoration
FormulaMathNeeds alt text that reads the formula aloud
LinkA linkHolds the link text and the clickable link area (an annotation)
Form, AnnotForm fields and other annotationsEach fillable field sits in a Form tag
CaptionA caption for a table or figureSits next to the item it describes
TOC, TOCITable of contents and its entriesEntries usually contain links
SpanA piece of text inside a paragraphOften used to mark a phrase in another language
Note, ReferenceFootnotes or endnotes, and references to them

Our guides on tables, forms and alt text go deeper on those tags.

Artifacts: content screen readers should skip

"Artifact" isn't a tag. It's how a PDF marks content that isn't part of the document's meaning: running headers and footers, page numbers, decorative lines, background shapes and watermarks. Artifacts sit outside the tag tree, so assistive technology skips them, and readers don't hear "Page 4 of 12" in the middle of a sentence.

Under PDF/UA, every piece of content must be either tagged or marked as an artifact. Nothing can be left in between. When a checker reports content is tagged or artifacted, it has found content that's neither.

How to decide:

  • Page numbers, running headers and footers: artifact.
  • Decorative borders, lines and background images: artifact.
  • A logo next to the organization's name in text: usually artifact.
  • A photo or chart that adds information: Figure with alt text.

Role maps: custom tag names

Word, InDesign and other tools sometimes create tags named after styles, such as BodyText or Subtitle. Assistive technology only understands the standard types, so the PDF carries a role map that says what each custom tag means, for example BodyText means P.

Problems appear when a custom tag has no mapping, when a standard tag is remapped to something else (PDF/UA doesn't allow it), or when mappings loop back on themselves. PAC reports these as non-standard structure type is remapped. In InDesign, you can avoid most of them by mapping each paragraph style to a standard tag before export.

Tagged is not the same as accessible

Automatic tagging gets you a tag tree quickly, but it often gets the details wrong. Common problems in auto-tagged files:

  • Large text tagged P instead of a heading, or headings at the wrong level.
  • Tables tagged as paragraphs, or header rows tagged TD.
  • Page numbers and running headers tagged as content instead of artifacts.
  • Decorative images tagged as Figure with no alt text.
  • Columns read in the wrong order.
  • Lists tagged as paragraphs with typed bullets.

A checker that says "Tagged PDF: Passed" only confirms that tags exist. Whether they're right takes a closer look, partly by software and partly by a person.

How to tag a PDF for accessibility

  1. Tag at the source if you can. Export from Word or PowerPoint with Document structure tags for accessibility turned on, or from InDesign with Create Tagged PDF. Fixes you make in the source file survive the next export; fixes made in the PDF don't.
  2. If you only have the PDF, autotag it. In Adobe Acrobat Pro, use All tools > Prepare for accessibility > Automatically tag PDF (in older versions, Tools > Accessibility > Autotag Document). If the file already has tags, decide first whether to repair them or replace them, because autotagging can overwrite someone's earlier work.
  3. Walk the tag tree. Open the Tags panel and go top to bottom. Change wrong tag types (right-click the tag and choose Properties), drag tags into the right order, and delete empty ones.
  4. Use the Reading Order tool for bigger fixes. Choose All tools > Prepare for accessibility > Fix reading order, draw a rectangle around content and set it as a heading, figure, table or background (artifact).
  5. Finish the details. Add alt text, set table header cells and scope, and set the document title and language.
  6. Check and listen. Run Acrobat's checker and PAC, then listen to the document with a screen reader or PAC's screen reader preview.

Our step-by-step guide on how to make a PDF accessible covers each of these in more detail.

How to view a PDF's tags

  • Adobe Acrobat Pro: open the Tags panel. In the current interface it's listed as Accessibility tags under View > Show/Hide > Side panels; in older versions it's View > Show/Hide > Navigation Panes > Tags. Choose Highlight Content from the panel's options menu to see which content each tag covers.
  • PAC: this free Windows checker tests PDF/UA and WCAG requirements, and its screen reader and structure preview shows what a screen reader will read, and in what order.
  • Includoc's free PDF tag viewer: upload a PDF and see its tag tree and reading order in your browser, with no software to install.
  • A screen reader: NVDA (free for Windows) or VoiceOver (built into Macs) shows you the result rather than the structure, which is what matters most.

The free Adobe Acrobat Reader can tell you whether a file is tagged (File > Properties), but it can't show or edit the tags.

Fix it automatically

Includoc builds a full tag tree for untagged PDFs, or repairs the existing tags when they're good enough to keep. It maps custom tags to standard ones, marks page furniture as artifacts, tags links, form fields and annotations, and sets a column-aware reading order. Our AI pass then checks heading levels, lists, table headers and reading order against the rendered page. Anything that needs judgment, such as alt text or a complex table, goes to a review screen for a person to confirm.

Upload your PDF to fix this automatically

Free check in seconds. Files are deleted within 24 hours.

Check your PDF

Standards

Standards references: This PDF has no tags
Standard or toolReference
WCAG 2.1
PDF/UA-1 (ISO 14289-1)Clause 7.1
Acrobat ruleTagged PDF
Standards references: Some content isn't tagged or marked as decoration
Standard or toolReference
WCAG 2.1
PDF/UA-1 (ISO 14289-1)Clause 7.1
Matterhorn ProtocolCheckpoint 01-005
PAC wordingContent is tagged or artifacted
Acrobat ruleTagged content
Standards references: Custom tags aren't mapped to standard ones
Standard or toolReference
WCAG 2.1
PDF/UA-1 (ISO 14289-1)Clause 7.1
Matterhorn ProtocolCheckpoint 02-001
PAC wordingNon-standard structure type is remapped

Sources

Related