A safe way to remove hyperlinks from a Word document is to first extract every external URL with a local browser tool, then use the resulting list to decide which links to delete, replace, or keep in the original .docx. The Word Hyperlink Extractor does this by reading the document's Office Open XML relationship file inside your browser, returning one clean http, https, or mailto address per line in a downloadable TXT report. Because the file never leaves your device, the audit can run on confidential reports, client templates, or internal memos without uploading them. The list deliberately excludes internal bookmarks, file paths, relative targets, and javascript-style URLs, so the count reflects safe external destinations rather than every clickable object Word might display. Anchor text and page position are not preserved in the report, which is why the original .docx stays the source of truth and the TXT file becomes a precise removal checklist. Treat each line as a document-supplied string that still deserves a quick review before any link is deleted from the body of the document.
Why Word Hyperlinks Are Harder to Find Than They Look
Word documents often hide useful destinations behind short phrases, button labels, citations, and even images. A reviewer who copies visible text to look for URLs will miss link targets attached to a single word, a citation marker, or a graphic. The same is true when links were inserted by mail merge, by a citation manager such as Word's built-in bibliography tools, or by an older Find and Replace pattern that bound text to a URL behind the scenes. These targets live in the document package rather than in the prose, which is why removing them requires more than a casual read-through.
The Office Open XML format used by modern .docx files stores external links in a sidecar file called the relationship part. According to the ECMA-376 Office Open XML standard, this part names each link, records its protocol, marks whether the target is external, and points to the destination address. Word itself can show the address when you hover, but it cannot export the entire list in one step.
This is the gap an extractor fills. By reading the relationship part instead of guessing from visible text, a focused tool can return the external hyperlink destinations recorded in the document's relationship file. The audit treats the relationship file as the source of truth and the visible prose as the surface.
What the Word Hyperlink Extractor Reads From a .docx
The extractor works against a single .docx file chosen from your device. Behind the scenes, every modern Word document is a ZIP package containing the main body, styles, and a folder of relationships. The relationship file relevant to hyperlinks is word/_rels/document.xml.rels, and it stores each link as a Relationship entry with a type of hyperlink, a target mode of External, and a target URL.
The tool reads only that relationship file. It validates the ZIP package first, then opens the XML, keeps entries whose type and target mode qualify as a safe external hyperlink, decodes the XML entities in the target string, normalizes the protocol to lowercase, and removes duplicates. The cleaned list is shown on the page and offered as a TXT download with one URL per line.
Because the extraction is local, your file stays in the browser. The relationship XML is read only after you select a .docx, the page does not crawl any destination, and the resulting TXT report contains only the URL strings. The Microsoft Open XML documentation describes this same relationship structure, confirming that the targets the extractor reads are the same addresses Word itself resolves when a user follows a link.
How to Extract Hyperlinks from a Word Document
The process is short, and every step happens in the browser tab. Before you start, close the original .docx in Word so the file is fully saved and unlocked on disk.
- Open the Word Hyperlink Extractor page and choose one .docx Word document from your device using the file picker.
- Wait while the browser reads the external hyperlink relationships from the document package. No upload happens; the relationship XML is parsed locally.
- Review the safe destinations shown in the result panel. Each line is one unique external URL extracted from word/_rels/document.xml.rels.
- Download the one-link-per-line TXT report and save it next to your original .docx for reference.
- Reopen the .docx in Word, compare the TXT list with the visible source, and remove or update each link you no longer want.
If the browser reports that the package is malformed, the relationship XML is oversized, or the archive uses unsupported ZIP64 directory records, the extraction stops before any URL is returned. Those local checks keep a simple extractor from pretending to understand a file it cannot safely read, which is the right boundary for a focused relationship reader.
What the Tool Includes and What It Skips
The extractor is deliberately conservative. The categories below come from the official scope of the relationship reader and apply to every document you load.
| Included in the result | Excluded by design |
|---|---|
| http and https destinations | Internal headings and bookmarks |
| mailto: addresses | File-system paths and relative targets |
| Targets marked with TargetMode=External | javascript: and other unsafe protocols |
| Unique destinations after deduplication | Duplicate destinations of the same URL |
| Targets that pass ZIP and XML validation | Malformed XML values and unsupported ZIP64 records |
Two consequences follow. First, the line count represents safe external destinations, not every clickable object Word might display in the document. Second, the report does not attempt to preserve the visible anchor text, its page location, formatting, or surrounding paragraph; those details remain in the original .docx and cannot be reconstructed from a list of bare URLs.
After the List: Choosing Which Links to Remove
With the TXT report in hand, the removal decision becomes a comparison exercise rather than a hunt. Open the .docx in Word, run Find and Replace or use the Remove Hyperlink command on each item you flagged, and let the original document carry the surrounding sentence for context. The TXT file is short enough to paste into a spreadsheet if you want to triage sources into keep, update, or delete buckets before editing.
Several built-in Word paths still apply after the audit. Right-clicking a single link and choosing Remove Hyperlink strips the URL but keeps the anchor text. Pressing Ctrl+A to select all body text, then pressing Ctrl+Shift+F9, clears all field codes including hyperlink bindings, which is the quickest way to remove every Word hyperlink at once once you have confirmed in the audit that nothing critical will be lost. Paste Special as Unformatted Text strips links from copied content before it lands in Word, which is useful when the document is still being assembled from outside sources.
The extractor does not perform any of these removals. It produces the audit, and the original .docx is still required for context, anchor text, and the actual hyperlink deletion step. Treat the relationship file as the inventory and the visible body as the editing surface, then act on one while reading from the other.
When an Audit Beats Clicking Through the Document
Reviewing a long report by hovering over each link is slow and easy to skip. Reviewing a TXT list of unique external destinations is fast and lets you check each one through whatever separate process your team uses for link safety. This is the workflow the extractor supports: collect, inspect, decide, then act on the original document.
The same approach fits other jobs. Auditing a template before it ships to a client, collecting references from a document received from another team, and building a concise handoff list for a new editor all start with the same question: which external URLs does this file actually point at. The extractor answers that question without guessing from visible text, without visiting any destination, and without sending the document away from your machine. No domain reputation is scored, no redirect is expanded, and no spreadsheet export is promised, because the scope is a precise list of safe-protocol external relationship targets and nothing more.
Keep the original .docx when you need anchor text or page context, download the TXT report as a precise list of safe external hyperlink relationships, and use the two together to plan every removal. The relationship file is the source of truth for where a Word document links; the visible text is only the surface.
If you're weighing options, How to Find and Replace Text in Illustrator Safely covers this in detail.