A colleague who uses a screen reader opens your PDF and hears the third paragraph before the first heading. An image caption is read aloud in the middle of a bulleted list where it does not belong. A sidebar about an unrelated topic interrupts the main text mid-sentence. The content is all there and the document looks fine on screen, but for anyone relying on assistive technology, the reading order is scrambled and the document is difficult or impossible to follow.
Reading order determines the sequence in which a screen reader reads the elements on a PDF page. When the reading order is correct, the experience is smooth and logical. When it is wrong, the content becomes a jumbled stream of disconnected fragments. The problem is invisible to sighted readers because they process the page visually, jumping between elements in whatever order catches their attention. For users who navigate linearly by keyboard, Tab key, or screen reader, the reading order is the only order.
Why PDF Reading Order Goes Wrong in the First Place
PDFs do not inherently understand reading order. When an application exports a document to PDF, it writes content objects to the file in whatever sequence the export engine processes them. This sequence is often the order in which elements were created in the source document, not the logical reading order a human would follow. A text box added to a page layout after the main body text may appear later in the object stream, even though it should be read first as a pull quote or callout.
Multi-column layouts are especially prone to reading order problems. The export engine may write all of column one followed by all of column two, which is correct, or it may write the first line of column one, then the first line of column two, then the second line of column one, and so on, which produces incomprehensible alternating text. The same issue affects documents with sidebars, callout boxes, image captions, footnotes, and headers or footers that contain substantive text.
Tagged PDFs are supposed to solve this. PDF Tags define a hierarchical structure, similar to HTML, that tells assistive technology the correct reading sequence, forming the backbone of PDF Accessibility for screen reader users. But tags are only as good as the software that generates them. A PDF exported from Microsoft Word with heading styles properly used will have a reasonably good tag structure. A PDF exported from a graphic design application where the designer manually positioned every text frame may have no tags at all or tags in an essentially random order. When the tag tree is absent or broken, the screen reader falls back to the content object order, which is often wrong.
How to Check the Reading Order of Your PDF
Adobe Acrobat Pro provides the most direct reading order inspection tools. Open the PDF and go to the Accessibility toolset, then select Reading Order. Acrobat displays numbered overlays on each page element showing the current reading sequence. Each number corresponds to the order in which a screen reader will encounter that element. Drag and drop to reorder elements if the sequence is wrong. The tool also shows the content type assigned to each element: paragraph, heading, figure, table, artifact. An element tagged as a paragraph that should be a heading will not appear in the screen reader's heading navigation list.
For a quick manual test that requires no special software, open the PDF in Acrobat Reader and use the Read Out Loud feature under the View menu. Activate it and listen. Do not watch the screen while it reads. Close your eyes if it helps. Does the narration follow a logical path through the content? Does it announce the title, then the first heading, then the body text under that heading? Or does it jump from a sentence in the introduction to a footnote, then to a caption on the third page, then back to the body text? The sequence you hear is exactly what a screen reader user experiences.
Common Reading Order Failures and Their Fixes
Headers and footers containing document titles, page numbers, or dates often interrupt the reading flow because they appear on every page and are read aloud each time, breaking the continuity between the last paragraph of one page and the first paragraph of the next. The fix is to mark header and footer content as artifacts in the tag tree. Artifacts are ignored by screen readers. This tells the assistive technology that the content is decorative or repetitive and should be skipped during linear reading.
Decorative images and background elements that have inadvertently been assigned alt text or that the screen reader treats as content objects cause another common interruption. A background gradient or a decorative border pattern has no informational value and should be marked as an artifact. In the Reading Order tool, right-click the element and select "Background" or "Decorative" to remove it from the reading sequence. A clean reading order includes only the content that a human would intentionally read, in the sequence they would read it.
Tables present a specific reading order challenge. A screen reader should read a table row by row, cell by cell, announcing the column header before each data cell so the listener knows what each value represents. If the table is not properly tagged with table header cells marked as TH and data cells marked as TD, the screen reader reads the table as a continuous run of text with no structural context. A data table that reads as "Q1 Q2 Q3 Q4 120 145 160 190" is meaningless. The same table read with header association becomes "Q1 120, Q2 145, Q3 160, Q4 190," which conveys the data correctly.
How to Fix the Reading Order Without Acrobat Pro
If you do not have access to Acrobat Pro's Reading Order tool, the most effective fix is to return to the source document and ensure the reading order is correct there before exporting to PDF. In Microsoft Word, use the Navigation Pane to verify that headings follow a logical hierarchy. Run the built-in Accessibility Checker under File, Info, Check for Issues, Check Accessibility. Fix heading levels, add alt text to images, and mark table header rows before exporting. A clean source document produces a clean PDF more reliably than trying to fix a broken PDF after export.
WukongPDF's editing tools can reorder pages and restructure content within a PDF Reading workflow by allowing you to rearrange pages into the correct sequence. While the browser-based editor does not provide the tag-level control that Acrobat Pro's Reading Order tool does, it handles the most common scenario: a PDF where entire pages are out of order because files were merged in the wrong sequence or scanned pages were inserted at incorrect positions. For full tag-level remediation, a dedicated accessibility tool is still the best option, but page-level reordering solves the problem that users encounter most often.
Preventing Reading Order Problems at the Creation Stage
The most efficient way to ensure correct reading order is to author with accessibility in mind from the beginning. Use heading styles, not manual font size and bold formatting, to create your document's structure. Place images inline with text rather than floating them in arbitrary positions. Use the application's built-in table, list, and column tools rather than simulating layouts with tabs and spaces. Each of these choices at the authoring stage directly determines the quality of the tag structure in the exported PDF.
When exporting to PDF, always select the option to create a tagged PDF and to export bookmarks and hyperlinks. These settings are typically found in the PDF export options dialog under an accessibility or structure heading. In Microsoft Word, the option is "Document structure tags for accessibility." In Adobe InDesign, it is "Create Tagged PDF." In Google Docs, tagging is automatic when you download as PDF. Checking that box takes one second. Fixing an untagged PDF after the fact can take hours. The small habits of structured authoring and correct export settings eliminate the reading order problems that frustrate assistive technology users and create compliance headaches.