SumatraPDF PDF Search Not Finding Text: How to Fix It

You can see a word in a PDF, yet SumatraPDF reports that it cannot find it. This usually means one of four things: the search command is not active in the expected document, the visible page contains an image rather than searchable text, the PDF stores characters in an unusual way, or a local SumatraPDF setting is not behaving correctly. The steps below focus specifically on in-document search with Ctrl + F or the slash shortcut, not web search or external selection-search services.

1. Confirm the Symptom With One Simple Test

Before changing settings, determine whether the problem follows SumatraPDF or the document. This avoids reinstalling a working application when the PDF itself has no usable text layer.

1.1 Search for a simple visible word

Open the affected document and choose a short, ordinary word that is clearly visible on the current page. Prefer a word containing only basic letters, such as “report,” “total,” or “chapter.” Avoid punctuation, accented characters, line breaks, and words at the end of a line for this first test.

1. Press Ctrl + F. You can also press / to open in-document search.
2. Type the selected word without quotation marks.
3. Press Enter or use the next-match control.
4. If no result appears, try a shorter portion of the same word.

SumatraPDF search is normally case-insensitive, so searching for “report” should also locate “Report” or “REPORT.” Changing capitalization is therefore unlikely to fix a failed search.

Success looks like this: SumatraPDF highlights the visible word or moves to a page containing it. If that happens, stop changing settings. The original query was probably affected by punctuation, line structure, hyphenation, or character encoding.

1.2 Test text selection and copying

Try selecting the same visible word with the mouse, then copy it and paste it into Notepad. This is one of the fastest ways to learn what the document actually contains.

• If the pasted text is readable and matches the page, the document probably has a usable text layer.
• If you cannot select individual words, the page may be a scanned image.
• If pasting produces blank text, random symbols, or different letters, the PDF may have a broken or unusual character map.
• If the pasted text contains unexpected spaces or line breaks, search may not match the phrase as displayed.

Copying is a diagnostic test, not a perfect guarantee. A PDF can sometimes support selection while still storing text in an order or encoding that makes phrase searches unreliable.

1.3 Compare a known-good PDF

Open another PDF that you know contains digitally generated text, such as a document exported from a word processor. Search for a visible word in that file.

If search works in the known-good PDF, SumatraPDF's search feature is functioning and the affected document is the likely cause. Stop changing global settings and concentrate on the source document. If search fails in every normal text PDF, continue with the application checks below.

2. Check the Search Command and Document Format

SumatraPDF reads several document types, but visible content is not always represented as searchable text. PDFs, eBooks, comics, CHM files, DjVu documents, XPS files, and images can store content differently. Search capability therefore depends partly on the format and how the file was created.

2.1 Make sure the search box belongs to the active document

Click inside the document you want to search before pressing Ctrl + F or /. If multiple windows or tabs are open, verify that the intended document is active. Enter the query again instead of relying on text left over from another search.

For a basic test, search forward from the beginning of the file. A search that begins after the only occurrence may need to wrap around before locating it. Watch for any message indicating that the end of the document was reached.

Success looks like this: the current file receives focus, the find interface appears, and navigating to the next result highlights an occurrence. Once this works, no file association or installation change is needed.

2.2 Account for hyphenated and split text

Text that looks continuous on the page may not be continuous internally. A word divided at the end of a line might be stored as “trouble-” followed by “shooting” on the next line. A heading may contain separate positioned text objects, even though it looks like one phrase.

Try these query variations:

• Search for only the first distinctive part of the word.
• Search for each word in a phrase separately.
• Remove punctuation and hyphens.
• Avoid copying a line break into the search field.
• Try a nearby plain word to confirm that search works elsewhere on the page.

If a shorter fragment is found, SumatraPDF is working. The limitation is the PDF's text layout, so resetting the application will not improve that document.

2.3 Consider ligatures and character encoding

Some PDFs store printed letter combinations as ligature characters. For example, “fi” or “fl” may be represented by a single glyph instead of two ordinary letters. Other files display the correct shapes while mapping those shapes to incorrect Unicode characters. This can cause a visible word to be missed even when it looks perfectly normal.

Copy the word from the PDF and paste it into Notepad. Compare the pasted result carefully. You can also paste the copied text directly into SumatraPDF's search box. If the pasted version works but typing the same-looking word does not, the document uses characters that differ internally from ordinary keyboard input.

This is a document-encoding issue rather than a case-sensitivity setting. The durable fix is to obtain a properly exported or OCR-processed copy from the document owner or source.

3. Check the Source Document and Windows Factors

3.1 Identify scanned PDFs without OCR

A scanned PDF often contains one image per page. The image shows letters to a human reader, but the file contains no actual words for SumatraPDF to search. SumatraPDF is a reader, not an automatic OCR conversion tool.

Signs of an image-only scan include:

• Dragging selects a rectangular area or nothing at all instead of individual characters.
• Copied content is blank or unavailable.
• Every word fails, including large, simple headings.
• Another PDF viewer also cannot find visible words.
• Extreme zoom reveals letters that look like pixels in a photograph.

To make such a file searchable, use a legitimate OCR tool or request an OCR-enabled copy from the sender. OCR creates a text layer associated with the scanned page. Do not attempt to bypass document permissions or access restrictions.

Success looks like this: after authorized OCR processing, individual words can be selected, copied as readable text, and found with Ctrl + F. If another viewer confirms that the original has no searchable text, stop modifying SumatraPDF.

3.2 Try another viewer as a control

Open the same PDF in another reputable viewer or in a browser's built-in PDF reader. Search for the identical simple word.

• If no viewer finds it, the PDF lacks searchable text or has defective text encoding.
• If another viewer finds it but SumatraPDF does not, update SumatraPDF from its official website and repeat the test.
• If only complex phrases fail, compare how each viewer handles line breaks, reading order, and ligatures.

This comparison is not a recommendation to replace SumatraPDF. It is a controlled test that separates a file problem from an application-specific parsing problem.

3.3 Verify which SumatraPDF copy Windows opens

Windows file associations can open a different SumatraPDF executable from the one you intended to test. This is especially possible if both installed and portable copies exist.

1. Open SumatraPDF directly from the Start menu or the portable folder you intend to use.
2. Use the application's About information to identify the running copy.
3. Open the affected PDF from inside that application.
4. Repeat the simple-word search.

If search works when launching one copy directly but not when double-clicking PDFs, the file association is pointing to another installation. Set the desired application as the default PDF reader through Windows settings, then test again.

3.4 Know which external components are relevant

For ordinary PDF text search, printer drivers, print settings, and Windows Search indexing are not involved. Windows Search controls operating-system file searches, not Ctrl + F inside an open SumatraPDF document. Likewise, changing the default printer will not create a missing PDF text layer.

Ghostscript may matter in workflows involving PostScript files or conversions, but it is not normally required to search standard PDF text in SumatraPDF. Image codecs can affect whether certain image formats display, but they cannot turn letters inside a scanned image into searchable text. Only investigate these components when the original format or conversion workflow specifically depends on them.

4. Inspect or Reset SumatraPDF-settings.txt Safely

If search fails in multiple known-good text PDFs, a clean settings test is reasonable. SumatraPDF stores user preferences in a text file named SumatraPDF-settings.txt. Installed and portable copies can use different settings locations, which explains why behavior may differ between them.

4.1 Back up the settings file first

Close every SumatraPDF window before changing the file. For an installed copy, the settings file is commonly stored under the current user's local application data in a SumatraPDF folder. A portable copy may keep its settings with the executable when that location is writable.

Do not edit unfamiliar values at random. Instead:

1. Locate SumatraPDF-settings.txt.
2. Copy it to a safe backup location.
3. Rename the original to something such as SumatraPDF-settings-backup.txt.
4. Start SumatraPDF again so it can create fresh defaults.
5. Open a known-good PDF and test Ctrl + F.

Success looks like this: search works with the fresh settings file. At that point, keep the new file and reapply only the preferences you need through the application. Avoid copying the entire old file back, because that may restore the problem.

4.2 Restore settings when the reset changes nothing

If search still fails, close SumatraPDF, remove the newly generated settings file, and rename your backup to its original name. This restores your preferences and confirms that settings persistence was not the cause.

A settings reset cannot add OCR, repair a malformed character map, or join words that the document stores as separate objects. Stop resetting settings when known-good PDFs search correctly or when another viewer proves that the affected file has no searchable text.

5. Run a Clean Temporary Test Before Reinstalling

Reinstallation is often unnecessary because user settings may survive an uninstall, while document-specific text problems remain unchanged. A cleaner diagnostic approach is to test a fresh portable copy without disturbing the installed reader.

1. Download SumatraPDF only from its official website.
2. Choose the portable package and extract it into a new temporary folder.
3. Do not copy your existing settings file into that folder.
4. Launch the portable executable directly.
5. Open a known-good, digitally generated PDF.
6. Search for a simple visible word using Ctrl + F and then the slash shortcut.
7. Finally, test the original problem document.

If the known-good file works but the original does not, the source PDF is responsible. If both work in the clean portable copy, the installed copy or its settings environment is responsible. You can then reset the installed settings or replace the installation with a current official build.

If neither file works in the clean test, verify the known-good file in another viewer and confirm that you are entering the query into the find box. A reproducible failure involving normal searchable PDFs may justify reporting the issue to the SumatraPDF project with a non-confidential sample file and precise reproduction steps.

6. Quick Fix Checklist

• Click the active document and open find with Ctrl + F or /.
• Search for a short, plain word visible on the current page.
• Do not troubleshoot capitalization because normal search is case-insensitive.
• Try part of a hyphenated word and search phrase words separately.
• Copy the visible text into Notepad and inspect what is actually extracted.
• Test a known-good PDF containing digitally generated text.
• Open the problem file in another viewer to confirm whether its text is searchable.
• Use authorized OCR when the PDF contains scanned page images without a text layer.
• Launch the intended installed or portable SumatraPDF copy directly.
• Back up and temporarily reset SumatraPDF-settings.txt only if all searchable PDFs fail.
• Run a fresh portable test before reinstalling or deleting preferences permanently.
• Stop changing settings once the evidence identifies a document-level problem.

7. Frequently Asked Questions

7.1 Why can I see text that SumatraPDF cannot find?

The visible letters may be pixels in a scanned image rather than text characters. Alternatively, the PDF may use unusual character mappings, ligatures, separate text objects, or an unexpected reading order. A viewer can render the page correctly while being unable to match the displayed word to ordinary keyboard input.

7.2 Does SumatraPDF search require exact capitalization?

No. In-document search is normally case-insensitive, so capitalization should not determine whether a standard word is found. Focus instead on spelling, punctuation, hyphenation, line breaks, ligatures, and whether the page has a real text layer.

7.3 What is the keyboard shortcut for searching inside a document?

Use Ctrl + F to open the find interface. The / key is also a shortcut for starting in-document search. These commands search the open document rather than the web or Windows file index.

7.4 Why does copying text help diagnose search failures?

Copying reveals what characters the PDF exposes to software. If the pasted output is correct, search should generally work for simple words. If it is blank, garbled, reordered, or unexpectedly spaced, the PDF's text layer is missing or defective. Pasting the copied characters into the find box can also expose ligature or encoding differences.

7.5 Will reinstalling SumatraPDF make a scanned PDF searchable?

No. Reinstalling a reader does not add OCR text to image-only pages. Use an authorized OCR application or obtain a searchable copy. Reinstallation is worth considering only when search fails in multiple verified text PDFs and a clean portable test points to the installed application.

7.6 When should I stop troubleshooting SumatraPDF settings?

Stop when a known-good text PDF searches correctly, when another viewer also fails on the affected file, or when copy and paste proves that the page has no usable text layer. Those results identify a document problem. Further preference resets, printer changes, or Windows indexing adjustments will not repair the source text.