Heading Structure Checker

What the Heading Structure Checker Does

Paste the HTML of a page and you get its heading skeleton: each H1 to H6 in order, indented by level, so you can see the document outline at a glance. Beneath it, the tool lists any structural warnings, or tells you the structure looks good.

It checks the shape of your headings, not their wording. A clean outline — one H1, then H2s for sections and H3s beneath them, with no jumps — helps screen-reader users navigate and helps search engines understand how the page is organised.

How It Works

The tool scans the pasted HTML for heading tags, capturing each heading's level and its text (stripping any tags nested inside). It prints them as an indented list, then runs three checks: is there exactly one H1, and does any heading jump more than one level deeper than the one before it.

Going back up the hierarchy is allowed — an H4 followed by an H2 is fine, because that just closes a subsection. The warning is only for going down too fast, like an H2 straight to an H4, which leaves a gap in the outline.

What it checks & how

1. Extract the headings. Find every H1 to H6 tag in the pasted HTML and read its level and text.
2. Build the outline. List the headings in order, indented by level, with nested tags stripped from the text.
3. Check the H1. Warn if there is no H1, or if there is more than one.
4. Check for skipped levels. Warn if any heading goes more than one level deeper than the previous one.

What it flags

Going back up a level (H4 then H2) is not flagged — that simply closes a subsection and is correct.

Technical Details

Standards & references

• One H1, logical order — the W3C guidance is one H1 for the page topic, then H2 for sections and H3 for subsections, without skipping ranks; style headings with CSS rather than picking a level for its size.
• WCAG headings — WCAG 1.3.1 (Level A) requires headings to be marked up semantically; 2.4.10 (AAA) calls for section headings to organise content. Screen-reader users jump between headings to navigate.
• Multiple H1s — the nuance — Google's John Mueller says multiple H1s are fine for search, but browsers never implemented the HTML5 outline algorithm, so screen readers treat every H1 as top-level — one H1 stays the clearer, more accessible choice.

Accuracy & Limitations

It reads the HTML with a regular expression, not a full browser parser. Clean, well-formed heading tags are detected reliably, but unusual markup — a heading whose opening tag contains a stray angle-bracket character inside an attribute, or broken nesting — can be missed or misread. Paste the real rendered HTML for best results.

It checks structure, not substance. It will not tell you whether a heading is descriptive, keyword-relevant or matches the section below it — only whether the H1 to H6 skeleton is sound.

The skipped-level warning fires only for downward jumps of more than one rank, like H2 to H4. Going back up (H4 to H2) is correct and is not flagged, since that closes a subsection.

Multiple H1s are flagged as a warning, but this is an accessibility and clarity nudge, not a Google error — Google tolerates several H1s. The warning reflects that screen readers, lacking the never-implemented HTML5 outline, treat every H1 as equally top-level.

It analyses whatever HTML you paste, not the live page. If your headings are added by JavaScript after load, copy the rendered HTML from the browser's inspector rather than the raw source, or they will be missing.

Real-World Use Cases

When to use it — and when not to

Need the headings from a live URL automatically? A crawler or browser extension reads the rendered page for you. Want a full accessibility audit? A tool like axe or WAVE checks contrast, labels and ARIA too.

Frequently Asked Questions

References

• Headings (W3C Web Accessibility Initiative)
• Understanding WCAG 1.3.1: Info and Relationships (W3C)
• Google Says H1 Headings Are Useful but Not Critical (Search Engine Journal)

Related Tools