DOCX to Markdown without breaking nested and numbered lists

Word outlines look neat in the ribbon and fall apart the moment you need plain text. Nested bullets become flat paragraphs. Numbered steps restart at 1. on every line. Tab-and-symbol “lists” that never used Word’s list styles turn into noise. If you convert DOCX to Markdown for docs, runbooks, or a knowledge base, list fidelity is the first QA gate. Use Word to Markdown on a real .docx, open the .md in a plain editor, and repair nests and numbering before you polish tone. This guide stays on lists. For the GitHub README and docs-PR loop, see Word to Markdown for GitHub READMEs and docs repos. For Obsidian vaults, see Import PDFs into Obsidian as Markdown.

Why Word lists break after DOCX to Markdown

Word stores list structure in styles, numbering definitions, and indent levels. Markdown stores the same idea as characters: - or *, 1., and spaces for nesting. Converters map what they can read. They do not invent hierarchy when the source used bold symbols, pasted bullets from email, or manual tabs.

Multi-level lists are the usual failure. Level-2 and level-3 items flatten to the same indent as level-1. Numbered sequences that restart in Word (section A steps 1-3, section B steps 1-4) may emit as one long 1. stream or as paragraphs with leftover digits. Mixed markers (bullet then number then bullet) confuse extractors that expect one list type per block.

Authors also fake lists. A line that starts with • typed from the keyboard, or with → and a tab, is body text with a decoration. It may survive as a weird character, not as a Markdown list item. Tracked changes and comment balloons around list items add more junk when someone converts without accepting revisions first.

You care about this when the list is the product: install steps, escalation trees, checklist acceptance criteria, legal outline clauses. A pretty paragraph dump fails the job even when every word is present. Treat list structure as content, not formatting garnish.

What list fidelity looks like in Markdown

Good Markdown lists are boring and consistent. Unordered items use - or * (pick one style for the file). Ordered items use 1. on every line or true incrementing numbers; most renderers renumber either way. Nested items indent by two or four spaces under the parent. Blank lines between items are optional; blank lines that split one list into many short lists are the bug to avoid.

A nested procedure should read like this in the source: parent step, then indented children, then the next parent at the outer indent. If you need a paragraph under a step, indent that paragraph to match the item, or break the step into shorter items. Do not leave a hanging sentence at column zero in the middle of a nest; many previewers will close the list early.

Numbered lists that document real sequence (reboot, wait, verify) must keep order in the rendered view. If conversion restarted every item at 1., rebuild the block. If Word used “continue numbering” across headings and Markdown lost the continuation, decide whether each section should restart. Restart is fine when each heading owns its own procedure. Silent merge of two procedures is not.

Task lists (- [ ]) are optional GFM. Convert them only when your destination supports them. Do not invent checkboxes Word never had. Keep definition lists and tables out of this pass unless a list cell depends on them; list fidelity first, table cleanup second.

Prepare the Word file so lists survive

Ten minutes in Word beats an hour of Markdown archaeology. Fix the source outline while you can still see levels in the navigation pane.

Open the document and select a multi-level list. Confirm the ribbon shows List Library or Multilevel List, not only Bold and a random glyph. Promote and demote with Tab / Shift+Tab so levels are real. Replace hand-typed • lines with Apply Bullets. Replace 1) typed by hand with a true numbered list when the sequence matters.

Accept or reject tracked changes before export when policy allows. Mixed revision text inside list items produces duplicate bullets or orphan numbers. Remove password protection. Prefer .docx over legacy .doc. Delete cover pages that only hold logos. Redact secrets: API keys, customer names, salary tables. Conversion copies readable text.

Real list styles vs fake bullets

Fake bullets look fine on screen and fail conversion. Select three “bullet” lines and check the style pane. If the style is Normal plus a symbol character, rebuild with Bullets. Same for numbered lines that are Normal with a typed 1.. Rebuild with Numbering so Word attaches a numbering definition the converter can walk.

Watch paste from chat apps and slide decks. Those pastes often bring non-breaking spaces and private-use bullet glyphs. Re-apply list formatting after paste. Then save.

Numbered procedures and restart traps

Decide restarts on purpose. After each Heading 2, either restart at 1 or continue from the previous section. In Word, right-click a number and use Restart at 1 or Continue Numbering. Document that choice in a comment for yourself; Markdown will not remember Word’s UI intent unless the numbering definition is clean.

Long handbooks with twenty mini-procedures should restart per section. One continuous 1-200 list across chapters is hard to maintain after conversion and hard to review. Split chapters into separate files before or after convert if the outline is that large.

Convert and inspect lists before you touch tone

Run conversion once on a representative chapter that contains nested bullets and a numbered procedure. Time the cleanup. Only then batch the rest. Scaling a broken nest pattern into thirty files wastes a week.

Keep Word open beside the editor while you work. You need the visual outline levels, not only the converted text. Mark which procedures are acceptance-critical so you spend fix time where a wrong nest would mislead someone.

Open Word to Markdown. Upload the prepared .docx under the size limit. Download the .md. Ignore voice and branding until lists and headings match the outline you care about.

Run DOCX to Markdown on a sample chapter

Pick the worst list chapter on purpose: install steps with sub-bullets, or an escalation tree. Convert that file alone. If the sample survives with light edits, convert sibling chapters the same way. If the sample flattens every nest, fix Word styles and convert again before you touch other files.

Save the download under a clear name (runbook-auth.md, not Document1.md). Keep the original .docx archived so you can re-convert after style fixes.

Spot flatten and restart failures

Scan for these patterns immediately:

• Items that should nest sit at the same indent as parents.
• Every ordered line shows 1. even when Word showed 1, 2, 3.
• Bullet characters (•, ○, ■) appear in the text instead of -.
• Extra blank lines split one procedure into many one-item lists.
• Heading text sits inside a list item because a list never closed.

Fix the first broken procedure fully. Use it as the template for the rest of the file. Do not sprinkle cosmetic edits across the document while structure is wrong.

Repair nested bullets and numbered lists by hand

Work in a plain editor with a Markdown preview pane (VS Code, Cursor, or a docs site preview). Rebuild one list at a time. Delete the broken block. Retype from the Word outline you can see, using - and consistent indent. For numbered steps, type 1. on each line or true numbers; then preview.

When a nest has three levels, keep level-3 rare. Deep nests are hard to read on mobile and in GitHub preview. Promote a deep branch to its own ### heading plus a short list when the branch is a real sub-procedure.

Mixed list types in one block (bullet, then number, then bullet) rarely survive cleanly. Split into two lists with a one-line lead-in sentence, or convert everything to bullets and move sequence into the wording (“First…”, “Then…”). Choose readability over clever nesting.

After repairs, search the file for •, ○, and leading tabs. Replace leftovers. Click through the preview once as if you were the on-call engineer following the steps cold. If you skip a nest level with your eye, a reader will too.

Related jobs: GitHub docs, PDF sources, notes, and RAG

List cleanup is shared work. Destination changes the success check.

Shipping a README or /docs page through a PR is a different loop: branch, preview on GitHub, review comments on lines. That workflow lives in Word to Markdown for GitHub READMEs and docs repos. Use this article’s list QA, then follow that guide for repo placement and review habits.

PDF-only packets need PDF to Markdown. Expect worse list fidelity than native .docx. Print-to-PDF often turns bullets into text boxes. If anyone still has Word, use that path instead. Mixed folders can use Convert to Markdown.

Personal notes in Obsidian belong on the vault path in Import PDFs into Obsidian as Markdown. You can drop the same cleaned .md into a vault later. Do not treat graph view as the QA for a runbook.

When the Markdown feeds search or an assistant, clean lists once so chunkers keep steps together. See Prepare documents for RAG with Markdown after the human-readable outline is stable.

Privacy, offline options, and when not to upload

Upload only files you are allowed to process in a third-party tool. Strip credentials first. Prefer a redacted excerpt when you only need one chapter online.

Offline conversion fits private corpora. Pandoc (pandoc file.docx -o file.md) and local CLIs run on your machine. Online tools win for one-off public drafts under size caps. Choose offline when the file holds secrets you refuse to upload. Browser-upload habits sit in Is it safe to upload documents online?. Retention and processing details are in the privacy policy.

Do not upload medical charts with identifiers, unreleased advisories, unsanitized contracts, or password lists. Convert those on hardware you control. Keep the Markdown private until policy clears publication.

Toolsy processes the upload for the conversion response, then drops it on the usual retention window. Free tiers cap daily files and size; split monster handbooks or use a plan that fits. None of that replaces your compliance rules.

Frequently asked questions

How do I convert DOCX to Markdown without losing nested lists?

Use real multilevel list styles in Word, then convert with Word to Markdown. Open the .md and compare nests to the Word outline before you edit tone. Rebuild any flattened block with - and consistent two- or four-space indent, then preview once as a cold reader.

Is docx to markdown different from word to markdown?

For modern Office files, no: .docx is what people mean when they search word to markdown. Legacy .doc is less reliable; reopen and save as .docx first. This article uses docx to markdown as the primary phrase because list QA targets the file format converters actually read.

Why do my numbered lists restart at 1 after conversion?

Word’s numbering definitions and “restart” settings often fail to map one-to-one into Markdown. Converters may emit 1. on every line or lose continuation across headings. Decide per section whether steps should restart, then rebuild the ordered list in the .md so the preview matches that decision.

Can a docx to markdown converter keep three-level bullet nests?

Sometimes, when Word used real multilevel lists. Fake bullets and paste junk usually flatten. After convert, check the deepest nest you care about. If level-3 vanished, retype it or promote that branch to a heading plus a shorter list.

Should I use Pandoc for pandoc docx to markdown instead of a browser tool?

Use Pandoc or another local CLI when the file is private, huge, or part of a scripted pipeline. Use an online converter for small drafts when policy allows upload. Many teams convert once in the browser to inspect list damage, then script Pandoc later for repeatable builds.

How do I convert a word document to markdown if I only have a PDF?

Prefer finding the .docx. If you only have PDF, use PDF to Markdown and budget more list cleanup. Print layouts often destroy indent. Treat the PDF path as last resort for procedures you must follow exactly.

Does Microsoft Word export Markdown that keeps lists intact?

Some Word builds and add-ins offer Markdown export; quality varies by version and list complexity. If export flattens nests, the same prep rules apply: real list styles, then a dedicated converter or Pandoc. Verify with preview; do not trust the first export blindly.

How do I keep lists intact for a GitHub README after conversion?

Fix nests and numbering here, then place the file in the repo and use GitHub Preview. The full PR and docs-site loop is covered in Word to Markdown for GitHub READMEs and docs repos. Do not merge raw downloads when on-call steps depend on indent.

Will this replace importing documents into Obsidian?

No. Vault import, naming, and note workflow belong in Import PDFs into Obsidian as Markdown. You may reuse a list-cleaned .md in Obsidian later. Success here is a faithful outline, not a graph view.

Is it safe to upload a company DOCX just to fix lists?

Only if policy allows and you redacted secrets. Read Is it safe to upload documents online? and the privacy policy. When the procedure is sensitive, convert offline and keep both Word and Markdown on systems you control.

For the convert step itself, start at Word to Markdown. When the only copy is PDF, use PDF to Markdown. Fix nested and numbered lists before you ship the file to a repo, a vault, or a search index.