Scanned PDFs often look perfectly organized visually but can be difficult or impossible for screen readers to understand correctly. Setting a logical reading order in a scanned PDF means making sure text, headings, images, tables, captions, and other content are presented to assistive technology in the sequence intended by the document author.
The process usually involves two stages: first, making the scanned pages searchable with OCR, and second, checking and correcting the document’s reading order and accessibility structure. In Adobe Acrobat, this can involve the Reading Order tool, tags, OCR, and accessibility checks. For complex documents, manually reviewing the tag structure is essential because automated tools can misinterpret columns, tables, sidebars, footnotes, and decorative elements.
Direct Answer: To set a logical reading order in a scanned PDF, run OCR to create accessible text, open the PDF’s accessibility or Reading Order tools, inspect the page sequence, and correct content blocks so a screen reader encounters headings, paragraphs, lists, tables, and images in the intended order. Then verify the result with a screen reader and accessibility checker.
What Does Logical Reading Order Mean in a PDF?
Logical reading order is the sequence in which assistive technologies, such as screen readers, interpret and present the content of a PDF.
A sighted reader may naturally understand a two-column page by looking from the top-left area across and then continuing down the page. A screen reader does not necessarily interpret visual positioning the same way. It relies on the PDF’s underlying text, tags, structure, and reading sequence.
For example, a properly structured article might follow this sequence:
Document title
Introduction
Heading
Paragraph
Subheading
Paragraph
Image
Image description
Table
Conclusion
If the underlying order is incorrect, a screen reader might announce the image before the heading, read the right column before the left column, or move through table cells in an unintuitive sequence.
Definition Box: Logical Reading Order
Logical reading order is the intended sequence of document content as interpreted by assistive technology. It determines whether users navigating with a screen reader encounter the PDF’s text, headings, images, lists, tables, and other elements in a meaningful order.
Logical reading order is related to, but not identical to, PDF tagging. Tags describe what content is, while reading order determines how that content is encountered.
Why Is Reading Order Important for Scanned PDFs?
A scanned PDF is fundamentally different from a digitally created PDF.
A digital PDF may contain actual text objects, headings, links, and structural tags. A scanned PDF commonly consists of page images. Without OCR, a screen reader may have little or no text to interpret.
Even after OCR, accessibility is not automatically guaranteed.
OCR can identify words while still producing problems such as:
Incorrect text sequence
Missing headings
Broken paragraphs
Incorrect table recognition
Misidentified columns
Missing alternative text
Incorrect language settings
Decorative content being treated as meaningful
Headers and footers being inserted into the middle of sentences
The following workflow is suitable for scanned documents being remediated for accessibility, particularly when using Adobe Acrobat.
Step 1: Create a Backup of the Original PDF
Before making accessibility changes, save an untouched copy of the original document.
Create a working copy with a descriptive filename such as:
Annual-Report-accessibility-remediation.pdf
This matters because OCR, tagging, and reading-order corrections can alter the document structure. Keeping the original allows you to compare results or restart the remediation process if necessary.
Step 2: Run OCR on the Scanned PDF
A scanned page is normally an image rather than selectable text.
For a two-column article, the intended sequence might be:
Title → Left-column content → Right-column content
or another sequence dictated by the document’s actual design.
Do not rely solely on the visual position of objects. Consider the document’s meaning and how a user should navigate it with a screen reader.
Step 6: Use the Reading Order Tool to Inspect Content Blocks
In Acrobat’s Reading Order interface, content may be identified as different regions, including text, images, tables, and other page elements.
Review the boundaries around each content block.
Look for common problems such as:
One paragraph split into multiple unrelated blocks
Two paragraphs merged together
A heading classified as ordinary text
An image classified as text
A table classified as several unrelated text blocks
Sidebar content inserted into the main article
Header and footer content appearing between paragraphs
The objective is to make the content sequence match the document’s intended meaning.
Step 7: Correct the Order of Multi-Column Content
Multi-column PDFs are among the most common reading-order problems.
Consider a page with two columns:
Visual Layout
Correct Screen Reader Sequence
Left column + right column
Left column → Right column
Main article + sidebar
Main article → relevant sidebar position
Heading spanning page + two columns
Heading → left column → right column
Footnote below article
Main content → footnote
The correct sequence depends on the document’s editorial structure.
Example
Suppose a page visually contains:
Heading
Left column:
Paragraph 1
Paragraph 2
Right column:
Paragraph 3
Paragraph 4
If Acrobat reads Paragraph 1 → Paragraph 3 → Paragraph 2 → Paragraph 4, the reading order is likely incorrect.
A screen reader user should encounter the content in the same logical progression intended by the document.
Step 8: Make Sure Headings Are Identified Correctly
A heading should not merely look bold or large.
For accessibility, headings should be represented using appropriate structural tags, such as:
<H1> for the primary document heading
<H2> for major sections
<H3> for subsections
A logical heading hierarchy helps screen reader users navigate long documents.
For example:
H1 — Employee Accessibility Guide
H2 — Workplace Accommodations
H3 — Requesting an Accommodation
H3 — Reviewing an Accommodation
H2 — Accessibility Training
Avoid choosing heading levels solely because they make text appear visually attractive. Heading structure should reflect the document hierarchy.
Step 9: Check Lists Carefully
OCR frequently creates problems with lists.
A visually correct list might become a series of unrelated paragraphs after OCR.
For example:
Open the document.
Run OCR.
Review the tags.
Test with a screen reader.
The accessibility structure should communicate that these items belong to a list.
Check whether:
List items remain in the correct sequence.
Bullets or numbers are not read as unrelated text.
Each list item is associated with the appropriate list structure.
Nested lists preserve their hierarchy.
Step 10: Fix Tables Separately
Tables require special attention because reading order alone cannot guarantee that the table makes sense.
A screen reader user needs to understand relationships between:
Column headers
Row headers
Data cells
Table captions where applicable
Example
Incorrect Structure
Accessible Goal
Table read as random text
Table structure preserved
Headers treated as body text
Headers identified
Cells read out of sequence
Logical cell navigation
Decorative borders interpreted as content
Decorative elements ignored
For complex tables, automated PDF remediation tools may not correctly determine the relationships between headers and cells. Manual tagging and testing may be required.
Step 11: Separate Decorative Elements From Meaningful Content
Not everything visible on a PDF page needs to be announced by a screen reader.
Decorative elements may include:
Background shapes
Decorative lines
Visual borders
Non-informative icons
Background images
If an element does not communicate meaningful information, it should generally not interrupt the reading experience.
Conversely, an image that communicates information should have an appropriate alternative text description.
Step 12: Add Alternative Text to Informative Images
Logical reading order and alternative text solve different accessibility problems.
Reading order answers: “Where should this content be encountered?”
Alternative text answers: “What does this meaningful image communicate?”
For example, a chart showing quarterly revenue should not simply be labeled “chart.”
A useful description should communicate the information a user needs from the image.
However, do not duplicate surrounding text unnecessarily. Alternative text should be concise when the surrounding content already explains the image.
Step 13: Check Headers and Footers
Headers and footers can create serious reading-order problems in long scanned documents.
For example, a screen reader could read:
“Accessibility Guide — Page 4 — Employees should request…”
instead of simply reading the sentence.
Repeated headers and footers should not unnecessarily interrupt the main narrative.
Review whether these elements are correctly identified and whether they should be treated as artifacts or meaningful content.
Step 14: Review the PDF Tag Tree
The tag tree provides a deeper structural view of the document.
A simplified structure might look like:
<Document>
<H1>
<P>
<H2>
<P>
<L>
<LI>
<LI>
<Figure>
<Table>
Review the hierarchy for:
Incorrect nesting
Missing structural tags
Duplicate content
Incorrect heading levels
Misidentified figures
Incorrect table structures
Out-of-order elements
For complex accessibility remediation, tag-tree inspection is often more reliable than visual inspection alone.
Step 15: Run an Accessibility Check
After correcting reading order, run an accessibility checker.
Adobe Acrobat includes an Accessibility Checker that can identify many common PDF accessibility issues.
Treat the checker as a diagnostic tool—not proof that the document is fully accessible.
Automated testing can detect many technical failures, but it cannot reliably determine whether every document is genuinely understandable to a screen reader user.
Step 16: Test With a Real Screen Reader
This is one of the most important steps.
Use a screen reader such as:
NVDA
JAWS
VoiceOver
Narrator
Navigate through representative pages.
Listen for:
Correct heading sequence
Logical paragraph order
Correct list navigation
Meaningful image descriptions
Sensible table navigation
Proper link announcements
Missing or duplicated content
A PDF can pass automated checks while still producing a confusing reading experience.
Common Reading-Order Problems and Their Fixes
Problem
Likely Cause
Recommended Fix
Columns read incorrectly
OCR or tagging sequence
Correct content order
Heading read as paragraph
Incorrect tag
Assign appropriate heading tag
Image announced incorrectly
Figure not tagged properly
Correct figure structure and alt text
Table read as random text
Poor OCR/tagging
Rebuild table structure
Header repeats throughout content
Header treated as normal content
Mark appropriately or artifact it
List read as separate paragraphs
Incorrect list structure
Apply list tags
Content missing
OCR failure
Re-run OCR or manually correct text
Decorative image announced
Image treated as meaningful
Mark as artifact where appropriate
What Is the Difference Between OCR, Reading Order, and PDF Tags?
These three concepts are often confused.
Feature
Main Purpose
OCR
Converts scanned image content into machine-readable text
Reading Order
Determines the sequence in which content is encountered
A scanned PDF can therefore have excellent OCR and still have an incorrect reading order.
Likewise, a document can have tags but still contain incorrect semantic relationships or poor alternative text.
How Do You Know If the Reading Order Is Correct?
A practical test is to temporarily ignore the visual design and experience the PDF as a nonvisual user would.
Ask:
Does the document title come first?
Are headings encountered before the sections they introduce?
Are paragraphs read in the intended sequence?
Are columns handled logically?
Are lists read as lists?
Are tables understandable?
Are informative images described?
Are decorative elements skipped?
Are headers and footers prevented from disrupting the narrative?
Can the document be navigated efficiently using headings and other structural features?
If the answer to these questions is yes, the document is much closer to a usable accessible PDF.
Accessibility Standards to Consider
For organizations working toward PDF accessibility compliance, WCAG 2.2 and PDF/UA are important reference points.
WCAG is a broader web accessibility standard, while PDF/UA specifically addresses accessibility requirements for PDF documents.
The appropriate compliance requirements depend on the organization’s legal obligations, jurisdiction, document purpose, and applicable accessibility standards.
Adobe’s official Acrobat accessibility documentation can also be used for product-specific remediation instructions.
Best Practices for Scanned PDF Accessibility
Use this checklist before publishing:
OCR
Text is searchable and selectable.
OCR errors have been corrected.
Important text has not been omitted.
Reading Order
Content follows a logical sequence.
Multi-column pages are correctly ordered.
Headings appear before their associated content.
Sidebars and footnotes are handled appropriately.
Structure
Heading hierarchy is meaningful.
Lists are structurally identified.
Tables have appropriate structure.
Links are identifiable and usable.
Images
Meaningful images have appropriate alternative text.
Decorative elements do not interrupt navigation.
Testing
Accessibility Checker has been run.
Tag structure has been reviewed.
Keyboard navigation has been tested.
Screen reader testing has been performed.
Why Automated PDF Accessibility Tools Are Not Enough
Automated remediation can dramatically reduce manual work, but it cannot fully understand document meaning in every situation.
A machine can recognize a rectangle as a table, for example, without necessarily understanding which cells function as headers. Similarly, OCR can recognize words without understanding whether a block is a heading, caption, sidebar, or footnote.
The most reliable workflow combines automated detection, manual remediation, and assistive-technology testing.
When Should You Rebuild a Scanned PDF Instead?
Sometimes remediation becomes more difficult than recreating the document from an accessible source.
Consider rebuilding the document when:
OCR quality is extremely poor.
The original source document is available.
Tables are highly complex.
The PDF contains hundreds of structural errors.
The scan quality is low.
The document contains extensive handwritten annotations.
The existing tag structure is fundamentally broken.
If the original Word, InDesign, PowerPoint, or other source file is available, creating an accessible source document and exporting a properly tagged PDF may produce a better result than repairing a heavily damaged scan.
Final Quality-Control Checklist
Before delivering an accessible scanned PDF, verify:
Text: Is OCR accurate?
Order: Does a screen reader follow the intended sequence?
Headings: Can users navigate the document by heading?
Tables: Are header and cell relationships understandable?
Images: Does every meaningful image have appropriate alternative text?
Decorative content: Is unnecessary content excluded from the reading experience?
Tags: Does the tag tree reflect the document’s structure?
Links: Can users identify and activate links?
Testing: Has the PDF been tested with an actual screen reader?
A PDF should not be considered fully accessible merely because it passes an automated checker. Human testing remains an important part of accessibility quality assurance.
Frequently Asked Questions
Why does my screen reader read a PDF in the wrong order?
Incorrect reading order usually results from OCR or PDF structure that does not match the document’s visual and semantic layout. Multi-column pages, tables, sidebars, captions, and headers are common sources of errors.
Can Adobe Acrobat fix PDF reading order automatically?
Acrobat can identify and assist with reading-order remediation, but automated detection is not always correct. Complex layouts should be manually reviewed and tested with a screen reader after remediation.
How do I check whether a PDF has a logical reading order?
Review the PDF’s Reading Order and tag structure, then navigate through it using a screen reader. Check headings, paragraphs, lists, tables, images, columns, and repeated page elements for logical sequencing.
Is reading order required for PDF accessibility?
Yes, meaningful reading order is a fundamental part of making a PDF usable with assistive technology. A visually correct PDF can still be inaccessible if its underlying content is presented in a confusing or incorrect sequence.
Pingback: Step-by-Step: Setting Logical Reading Order in Scanned PDFs - AccessiComply
Pingback: How to write descriptive alt-text for complex Excel bar charts exported to PDF. - AccessiComply