Why are fonts missing or different in my PDF?
Fonts usually look different because the PDF does not contain the font data needed to reproduce the original text, so the viewer substitutes another typeface. Missing letters, boxes or blank spaces can also point to an incomplete font subset, damaged font data or a conversion problem; start by checking whether the change appears in one viewer or in every viewer, then use the appropriate PDF compatibility fix.
A PDF can contain the complete font, a subset containing only the characters used, or merely a reference to a font that is expected to be available elsewhere. That difference determines whether the document keeps its original appearance when it is opened, printed or edited on another device.
A PDF needs its own font data to reproduce the original text
Fonts are software resources, not just names attached to text. A font provides the shapes of letters and other characters, along with measurements that control character width, spacing and positioning. If a PDF carries the required font data, a viewer can reproduce those shapes without depending on the fonts installed on the reader’s computer.
If that data is absent, the viewer may look for a locally installed font with the same name. When it cannot find one, it may substitute a different font. The replacement can be close enough to read while still changing line endings, spacing, page breaks, emphasis or the appearance of individual characters.
Fully embedded fonts carry the complete typeface data
A fully embedded font includes the complete set of permitted font information in the PDF. This is the most flexible form of embedding because the file contains more than the characters already visible on its pages.
Full embedding is useful when a workflow may need access to additional characters from the same font. However, it does not automatically give every recipient permission or technical ability to edit the text. Font licences can restrict embedding or editing, and a PDF editor may still require the font to be installed on the computer before it can add or replace characters.
Full embedding can also make a PDF larger, particularly when a document uses several fonts or fonts with very large character sets. For that reason, many export tools use subsets instead.
Subset-embedded fonts carry only the characters already used
A subset contains the glyphs—the visible character shapes—that the document used when the PDF was created. If a report uses the letters and numbers needed for its existing text, the PDF can store those glyphs without including every other character available in the original typeface.
A valid subset should display and print the existing text correctly. The limitation becomes important when someone edits the PDF. If the subset contains the letters in “Annual Report” but not the characters needed for new wording, an editor cannot obtain those missing shapes from the subset. It may require the original font to be installed, create another subset, or replace the edited text with a fallback font.
This explains a common symptom: an existing paragraph looks correct, but newly typed text in the same paragraph looks different. The PDF had enough embedded data to render the old content, not necessarily enough to create the new content.
Unembedded fonts leave the viewer to find a replacement
An unembedded font is not necessarily a problem on the computer that created the PDF. The correct font may be installed there, so the file appears normal during export and review. The failure appears after the PDF moves to a device that does not have the same font or has a different version of it.
Font substitution can produce several visible results:
- The typeface changes but the text remains readable. The viewer found a fallback font with the required characters.
- Words become wider or narrower. The replacement font uses different character measurements, which may alter wrapping or alignment.
- Only unusual symbols disappear. The substitute can display ordinary letters but lacks the specialist, mathematical or decorative glyphs used in the document.
- Characters appear as boxes, bullets or blank spaces. The viewer cannot map the stored character information to a usable glyph.
- Text looks correct on screen but changes when printed. The screen viewer and print process may be using different font resources or rendering paths.
- Edited text no longer matches nearby text. The original font is unavailable to the editor, which uses a fallback for the new characters.
Font embedding prevents ordinary substitution during viewing and printing, but it does not correct every text problem. Incorrect character encoding, damaged PDF objects, unsupported scripts or text converted badly during export can produce similar symptoms even when a font is listed as embedded.
The failure pattern identifies the PDF, viewer or export problem
Do not begin by installing random fonts or repeatedly saving the PDF. First establish where the change happens. A controlled comparison can separate a viewer-specific display issue from a fault already stored in the document.
Compare the PDF in more than one viewer
Open the same file in a dedicated desktop PDF reader and in a different viewer, such as another desktop application or a browser. Keep the zoom level similar and compare the same words, symbols and page positions.
The results narrow the diagnosis:
| What you observe | Most likely interpretation | What to check next |
|---|---|---|
| The text is wrong in one viewer but correct in another | The PDF contains usable information, but one viewer is substituting or rendering it differently | Update or change the affected viewer and test its font or language support |
| The same text is wrong in every viewer | The fault was probably written into the PDF during creation, conversion or editing | Inspect the font list and return to the source document if possible |
| The original text is correct, but edited text is different | The embedded subset supports the existing glyphs, while the editor lacks the original font for new text | Check whether the font is installed and permitted for editing |
| Only certain symbols or languages are missing | The available font or subset lacks the required glyphs, or the text uses problematic character mapping | Identify the affected font and script, then test a font that supports those characters |
| The PDF looks correct on screen but prints differently | The print workflow may be substituting fonts or interpreting the file differently | Print from another PDF application and inspect the printer or PDF-creation settings |
| Text has become fixed shapes and cannot be selected normally | The creator may have converted the lettering to outlines rather than preserving it as text | Use the source document to restore real text if search, editing or accessibility matters |
A browser rendering difference does not automatically prove that the browser is defective. It may simply expose a dependency that another viewer is handling more successfully. The goal is to make the PDF self-contained enough that its appearance does not depend on one particular application.
Inspect the PDF’s font list
In Adobe Acrobat, open the document properties and select the Fonts tab. Other professional PDF applications may provide a similar document-information or preflight panel.
Review each font used by the PDF. You are looking for three broad states:
- A font identified as Embedded contains full embedded font data.
- A font identified as Embedded Subset contains the glyphs used when that subset was created.
- A font with no embedding label may depend on a locally installed font or substitution.
Do not stop after inspecting the first font. A page can combine several resources: one for body text, another for headings, another for symbols and additional subsets introduced by later edits. If only a heading, formula or icon is broken, inspect the font assigned to that content rather than assuming the main body font caused the problem.
If all relevant fonts are embedded but characters are still missing, the problem may involve the mapping between stored character codes and glyphs, a damaged subset, an export defect or an unsupported viewer feature. Re-embedding the same faulty data may not help. Compare the PDF with the source document and try a fresh export.
Check when the change first appeared
The stage at which the text changed is often more informative than the visible symptom.
Ask these questions in order:
- Did the source document display correctly before export? If it did not, fix the source font before creating another PDF.
- Did the first exported PDF look correct? If not, review the export method, preset and font warnings.
- Did the problem begin after combining, optimizing or editing PDFs? A later process may have introduced a new subset, replaced text or removed a needed resource.
- Does the recipient see a problem that you cannot reproduce? Test on a device without the original authoring fonts. This can reveal that your local installation was hiding a missing-font dependency.
- Is the issue limited to newly entered form or annotation text? Form fields, comments and added text can use resources that differ from the fonts used in the original page content.
Re-export from the source when possible and repair the PDF only when necessary
The most reliable fix is usually to create a new PDF from the original document with suitable font-embedding settings. Editing the finished PDF is useful when the source is unavailable or the problem affects only a small amount of text, but it cannot always reconstruct font information that was never included.
Re-export with font embedding enabled
Open the source document on a computer that has the intended fonts installed. Confirm that every page, symbol and language displays correctly before exporting.
Then create a new PDF using the application’s direct PDF export command where possible. Review the PDF or advanced export settings for options such as:
- Embed all fonts.
- Embed OpenType fonts.
- Subset embedded fonts.
- Warn or stop when embedding fails.
- Use a print-quality or standards-based PDF preset.
The exact wording varies by application. After export, reopen the new PDF and inspect its font list rather than assuming the setting worked. A font can still be omitted when the application cannot access it or its licence settings prohibit embedding.
If a professional printer or another organization will process the PDF, ask whether it requires a particular PDF standard or preset. PDF/X workflows, for example, are designed for dependable print exchange and require fonts to be embedded. Use the recipient’s requested specification instead of choosing settings solely by file size.
Replace restricted or unavailable fonts in the source
Some fonts contain permissions that prevent or limit embedding. An export application should not be expected to override those restrictions.
If embedding fails:
- Read the warning or export report to identify the exact font.
- Confirm that the font is properly installed and activated.
- Check whether you are using the intended font version rather than a duplicate with the same family name.
- Review the font licence or vendor information for embedding rights.
- Replace the font in the source document if the required use is not permitted.
- Export a new PDF and inspect its font list again.
Choose a replacement that contains every required character, especially for multilingual documents, mathematical notation and specialist symbols. Visual similarity alone is not enough if the replacement lacks part of the character set.
Converting text to outlines can preserve letter shapes in a narrow print workflow, but it is a last resort rather than a normal embedding fix. Outlined lettering is no longer ordinary editable text and may reduce searchability, text extraction and accessibility. Keep a source version with live text even if a print supplier specifically requests outlined artwork.
Repair or edit the PDF when the source file is unavailable
If you have only the PDF, first save a backup. Then decide whether the goal is faithful viewing or editable text.
For a small visible error, a PDF editor may let you replace the affected text with an installed font. Select a typeface with suitable glyph coverage, then check line wrapping, spacing, alignment and page breaks. A metrically different replacement can move text even when its design looks similar.
Editing may be limited when the original font is embedded but not installed on your system. An embedded font can provide enough information to display existing characters without giving the editor everything it needs to compose new ones. Installing a legitimately licensed copy of the font may restore editing support; otherwise, replace the affected text with an available font.
A professional preflight tool can inspect fonts and, in some cases, embed a font that is installed and permitted for embedding. This is useful when the PDF content is otherwise correct. It will not manufacture an unavailable font, bypass licensing restrictions or reliably repair text whose character mapping was already damaged.
After any repair, test the result as a recipient would:
- Reopen the saved file rather than judging only the live editing view.
- Check it in a second PDF viewer.
- Zoom in on the corrected text and inspect punctuation and symbols.
- Search for a word from the repaired area.
- Copy and paste a short passage to check whether the text remains meaningful.
- Print or create a print preview if printed output matters.
- Inspect the font list again to confirm that the intended resources are embedded.
The durable fix is not simply to make the PDF look right on your own computer. It is to remove unnecessary dependence on local fonts and confirm that the saved file carries the resources needed to display its text consistently elsewhere.
Source: Font embedding, subsetting and substitution guidance from Font embedding and substitution in PDFs with Distiller, document font inspection guidance from Document properties and metadata overview in Acrobat, editing and fallback-font limitations from How to edit or format text in PDFs using Adobe Acrobat, and font inspection and correction guidance from Analyzing documents with the Preflight tool, Adobe. Page last checked 2 October 2026.