Accessible PDF documents

What we check

We collect the PDF links on the home page that point to the website's own domain and download up to four of them, up to 10 MB each. For every document we check:

• Text present: does the document have text on the first pages, or is it a scan, a mere image? An image without text is a warning.
• Tags: is the PDF structured with tags, that is headings, paragraphs, lists and tables that a screen reader understands? If they are missing, that is a warning.
• Language: is the document language set? Without it there is a note.
• Title: does the document have a title in its properties? Without one there is a note.

Encrypted or damaged documents we report as unreadable. If there are no PDF links, the check does not apply. The automatic check recognises the basics; whether the tags are set correctly and images have alternative texts only shows in a check with a screen reader.

Why it matters

On many websites PDFs are the most important content: price lists, forms, instructions, menus, annual reports. For blind and visually impaired people they are often the biggest barrier.

• A scanned PDF is a blank sheet for a screen reader. It reads nothing, the search in the document finds nothing, and Google cannot read the content either.
• A PDF without tags is read as a single block of text: without headings one could skip, tables as a jumble of numbers, multi-column layouts in the wrong order.
• Without a language the screen reader pronounces German text with English pronunciation. Without a title the document is called "flyer_final_v3.pdf" in the tab and in the screen reader.

The accessibility guidelines apply to PDFs as to web pages. Authorities and public enterprises are obliged to follow them in most countries, in the EU since 2025 many companies too. Public tenders explicitly require accessible PDFs.

How to fix it

When creating

• Word: use the styles for headings and lists, give images an alternative text and export with "File › Save as › PDF › Options › Document structure tags for accessibility". Language and title come from the document properties.
• InDesign: work with paragraph styles and the Articles panel for the reading order, and export with "Create tagged PDF".
• Google Docs, LibreOffice: both export PDFs with tags if headings are formatted as headings.

Existing PDFs

• Scans: run text recognition (OCR), in Acrobat under "Scan & OCR". Better: export the original from Word or the layout program again.
• Adding tags: in Acrobat Pro under "Accessibility › Autotag", then check the reading order. For many documents a specialised agency pays off.
• Language and title: in Acrobat under "File › Properties", tab "Description" for the title and "Advanced" for the language. This takes seconds, even for old documents.

Or no PDF at all

Content that changes or that many people read on a smartphone is better placed on a web page: opening hours, prices, instructions, menus. A web page is accessible, searchable and adapts to the screen.

How to check it yourself

Open the PDF in Adobe Acrobat Reader and press Ctrl+F (Mac: Cmd+F). Do you find a word you see on the page? If not, it is a scan. Under "File › Properties" you see title, language and whether the PDF has tags. The free "PAC 2024" (PDF Accessibility Checker) by the foundation "Access for all" checks PDFs in depth.

Frequently asked questions

How do I recognise whether a PDF has tags?

In Adobe Acrobat Reader under "File › Properties › Description", "Tagged PDF" says Yes or No. There you also see title and language.

Do all old PDFs have to be reworked?

The documents many people need come first: price lists, forms, instructions, annual reports. Old archive documents can be marked as such. Often a normal web page is the better solution instead of a PDF.

Why do you check only four PDFs?

Every PDF is downloaded and analysed, which takes time. Four documents linked from the home page show how the organisation behind the website creates PDFs. The rules apply to all of them.