#1233: WG Revision: CSSPseudoElement

Visit on Github

Opened Jun 2, 2026

Specification

https://drafts.csswg.org/css-pseudo-4/#CSSPseudoElement-interface

Explainer

https://github.com/danielsakhapov/CSSPseudoElementDoc

Links

Feature 1:

The specification

Where and by whom is the work is being done?

Feedback so far

I'm mostly requesting this review as I was asked in Intent to Ship process here - https://groups.google.com/a/chromium.org/g/blink-dev/c/CB44320ip2E/m/s9JMCrqDBgAJ - not sure if it even needs reviewing or can be just FYI even? PTAL

You should also know that...

https://github.com/danielsakhapov/CSSPseudoElementDoc - has links to some open issues on CSSPseudoElement interface

<!-- Content below this is maintained by @w3c-tag-bot -->

Track conversations at https://tag-github-bot.w3.org/gh/w3ctag/design-reviews/1233

Discussions

Log in to see TAG-private discussions.

Discussed Jun 8, 2026 (See Github)

Xiaocheng: Brian proposed a comment. I agree with it so we can post it.

Heather: I'll reach out to Brian, we'll go forward with it if he's cool with it.

Discussed Jul 6, 2026 (See Github)
<skip>
Comment by @bkardell Jul 15, 2026 (See Github)

Thanks for submitting this. The particular change that CSSPseudoElement no longer inherits from EventTarget, instead participating in event handling as Event.pseudoTarget which is a CSSPseudoElement improves compatibility and therefore sounds reasonable on its own. The CSS Working Group's resolution to “use the CSSPseudoElements interface for all standardized tree-abiding pseudo elements” also seems like a reasonable general extension to CSSOM.

Our concerns are regarding the accessibility implications of making pseudo-elements interactable, especially when applied to those well established ones like ::marker.

While Event.pseudoTarget is a necessary building block, this on its own leads to design patterns that are not accessible by default. For example, in the event-driven interaction use cases, the pseudo-elements can react to mouse by listening to the click event, but not to keyboard. Making them keyboard focusable will be another major problem that must be solved

However, making existing pseudo-elements like ::marker keyboard focusable might break web compatibility. Existing AT might not be able to detect that a ::marker is interactable.

We agree that interactable pseudo-elements can lead to useful design patterns, but it requires a lot more work than this interface and have some concerns that that work should be done early enough.