PDFs: Did the PDF Preserve the Document?

A PDF is not automatically accessible because it contains selectable text or passed through an export command. An accessible PDF needs meaningful tags, a logical reading order, correct document information, usable links and forms, and accessible alternatives for non-text content.

The fastest reliable workflow usually begins in the source document. Repairing Word, PowerPoint, or another authoring file is easier than rebuilding structure in Acrobat—and the improved source can be reused.

The question that matters

Does the PDF preserve the source document’s structure and meaning, or only its visual appearance?

A Practical Workflow

Find the source.
Whenever possible, repair the original Word, PowerPoint, or layout file instead of beginning in the PDF.
Make the source accessible.
Use styles, real lists, table headers, descriptive links, text alternatives, sufficient contrast, and a logical reading order.
Run the source checker.
Correct findings before conversion.
Export with accessibility enabled.
Use an export or save-as-PDF workflow that creates document structure tags. Avoid Print to PDF when it discards structure.
Check the PDF itself.
Verify tags, reading order, title, language, headings, links, tables, images, forms, and searchable text.
Repair or replace.
Return to the source when possible. If a complex PDF cannot be made reliable, provide an accessible HTML or source-document alternative and seek remediation support.

OCR is only the beginning

Optical character recognition can turn a scanned page into searchable text. It does not automatically create headings, lists, table structure, reading order, alt text, or accessible form fields.

Recognize the Starting Point

PDF conditions and next steps
What you have	Best next step
The editable source file	Repair the source, run its checker, export with tags, and verify the PDF.
A PDF with selectable text and tags	Run an accessibility check and manually inspect the tags and reading order.
A scanned image PDF	Locate the source or run OCR, then build and verify the required structure.
A complicated form, table, or layout with no source	Seek specialist remediation support or provide an accessible replacement format.
A document whose main purpose is online reading	Consider replacing the PDF with an accessible Canvas or web page.

Quick Check

You have an accessible Word handout and need a PDF. What is the strongest workflow?

Use Print to PDF and assume the visible formatting will preserve accessibility.: Printing can flatten or discard document structure.
Run Word’s checker, export using the accessibility-enabled PDF option, and check the resulting PDF.: Correct: Source structure helps create the tags, and the PDF still needs its own verification.
Take screenshots of each page and combine them into a PDF.: That replaces structured text with images and creates a much harder remediation problem.

Create and Check the PDF

From Microsoft 365

Complete the accessible Word, Excel, or PowerPoint source and run Check Accessibility.
Choose File > Save As or Save a Copy and select PDF.
On Windows, confirm Document structure tags for accessibility is selected in the PDF options.
On current Microsoft 365 for Mac, choose Best for electronic distribution and accessibility when available.
Open and inspect the exported PDF.

Microsoft guidance for creating accessible PDFs (opens in a new tab)

From Google Workspace

Make the Google document or presentation accessible before downloading it. Then choose the PDF download option and inspect the result. Do not assume all heading, table, equation, and reading-order information survived conversion.

Download as PDF option in the Google Docs File menu — Downloading creates the PDF; it does not complete the accessibility review.

In Adobe Acrobat Pro

Set the document title and language.
If the PDF is scanned, run OCR and carefully review recognition errors.
Confirm the document contains tags.
Run Accessibility Check and review every result rather than treating “passed” as the end.
Inspect the tag tree and reading order. Confirm headings, paragraphs, lists, tables, links, figures, and artifacts are identified correctly.
Check form-field names, instructions, keyboard order, and error handling when applicable.
Save a working copy and retest after repairs.

Adobe workflow for creating accessible PDFs (opens in a new tab)

When You Only Have a Scan

Search for an accessible original before remediating.
Use OCR to create text, then correct recognition errors.
Add the full logical structure; OCR alone is not enough.
Check page order, columns, footnotes, forms, tables, and images carefully.
For a long or complex document, seek remediation support or replace it with an accessible source.

A PDF can preserve only the accessibility built into the original document.

Before You Publish

The source document was repaired and checked before conversion.
The PDF was created with document-structure tags enabled.
Scanned text has been OCR’d and recognition errors reviewed.
The PDF has a meaningful title and correct document language.
Tags and reading order match the visible meaning.
Headings, lists, links, tables, images, equations, and forms are correctly represented.
Keyboard navigation, zoom, reflow, and text selection have been reviewed.
An accessible alternative is provided when reliable PDF remediation is not practical.

Convert a PDF to a Canvas Page with UDOIT

UDOIT can convert a PDF stored in Canvas into a Canvas page. This can be a useful alternative when the material is primarily meant to be read online, especially when repairing the PDF itself is not practical.

Open UDOIT from the Canvas course navigation and let the course scan finish.
Open Review Files, locate the PDF, and select Review.
Next to Canvas Page, select Request.
When processing is complete, open the generated page and compare it carefully with the PDF.
Repair headings, lists, tables, links, image descriptions, reading order, OCR errors, and formatting as needed.
Run UDOIT and the Canvas accessibility checker on the new page, then complete a manual review in Student View.
Only after verification, decide whether to keep both versions or use Replace File with a Page to update links to the PDF throughout the course.

Conversion is a starting point

The result depends on the source PDF. UDOIT can recover readable text from scans and create a more accessible Canvas page, but source errors may carry into the conversion. The generated page still requires human review for accuracy, structure, image alternatives, and meaning.

Cidi Labs: Convert a PDF document to a Canvas page in UDOIT (opens in a new tab)

Connection

A PDF is a delivery format, not a repair strategy. Strong source files create stronger PDFs, make future updates easier, and give you an accessible alternative when conversion does not preserve everything.