The Missing Sequel to CSS3

If you’ve written CSS for any length of time, you’ve probably noticed something odd: we never got a CSS4. The language has evolved dramatically since 2009, yet “CSS3” remains the last formally recognized version in most people’s minds. New features ship continuously—container queries, color-mix(), cascade layers—but there’s no shared vocabulary for talking about what “modern CSS” actually means or when those features arrived.

That’s the problem the CSS-Next community group is trying to solve. Officially still registered with the W3C as the “CSS4 Community Group,” CSS-Next brings together CSS Working Group members, browser vendors, designers, and developers to figure out how to describe CSS’s evolution in a way that’s useful for everyone—from educators to hiring managers to working engineers.

Why “CSS3” Stopped Making Sense

Back in 2009, CSS3 was a single meaningful milestone. It introduced a wave of features—border-radius among them—that felt like a genuine leap forward. Since then, each CSS specification has been versioned individually, which means there’s no aggregate release number to point to. The W3C’s draft specifications are public, and CSSWG discussions happen in the open, but there’s no clear answer to a simple question: What shipped, and when?

This isn’t just an academic annoyance. The CSS3 brand is still everywhere: job postings ask for CSS3 experience, educational materials display CSS3 logos, and tutorials claim to teach “modern CSS” while using a decade-old badge. The signal becomes useless when that same badge is expected to cover everything from Flexbox to features that didn’t exist when CSS3 was released.

The push for a successor started in earnest in 2020, when Safari Evangelist Jen Simmons opened an issue in the CSSWG’s GitHub repository asking whether CSS4 should be defined at all. The response was massive—one of the largest in the repo’s history based on emoji reactions alone. Support came from prominent voices like Chris Coyier, Nicole Sullivan, and PPK, who argued that a named release would help push technologies forward and give educators and site owners something concrete to point to.

Eras, Not Versions

The CSS-Next group isn’t trying to recreate the old versioning system. Instead, it’s exploring the idea of defining eras—bundles of features that shipped roughly together. Labels like “CSS4” or “CSS5” would serve as meta-information about when a feature entered the language, not as a statement about browser support.

That distinction matters. The web platform’s Baseline project already tracks browser support for individual features, telling you whether something is ready to use in production. But Baseline says nothing about where a feature fits in the language’s history. A feature can be baseline-available and still be relatively new to CSS; conversely, a feature from the CSS3 era might achieve baseline status only years later. Both signals are useful, but they answer different questions.

An annual snapshot approach—capturing the feature set once per year—sounds reasonable at first, but it breaks down in practice. Some years see few or no notable additions, producing nearly identical snapshots; other years bring a flood of new capabilities. Comparing whole-language snapshots year over year would demand an unreasonable amount of mental effort to identify what actually changed. An era-based approach groups features by meaningful milestones in the language’s trajectory, which is far easier to digest.

What’s at Stake

Without a successor to CSS3, the terminology dead-ends. The language has moved far beyond what anyone imagined in 2009—container queries alone would have seemed impossible to many—yet “CSS3” remains the only widely recognized milestone. That’s why the issue strikes a nerve with developers who see the same badge used to signify wildly different levels of knowledge.

The CSS-Next group is open to anyone who wants to participate. The goal is straightforward: find a way to talk about CSS’s evolution that doesn’t require developers to memorize a scatter of spec numbers, and give the community a shared reference point for years to come. The conversations are happening now, and the outcome will shape how the next generation of CSS practitioners learns and communicates about the language.

What CSS-Next Is — And Isn’t — Trying To Do

The CSS-Next group has rallied around a clear mission: give developers, educators, and employers a better way to talk about how CSS has changed since CSS3. The goal is not technical specification work or official documentation, but rather a shared framing that helps people learn, teach, and understand the language’s progression over time.

Specifically, the group wants to:

  • Help developers learn CSS.
  • Help educators teach CSS.
  • Help employers define modern web skills.
  • Help the community understand how CSS capabilities have evolved.
  • Create a shared vernacular for describing that evolution.

At the same time, the group explicitly does not intend to:

  • Affect spec definitions or influence working groups like the CSSWG.
  • Create official developer documentation, such as a new version of MDN.
  • Define browser specification work, which belongs in forums like the CSSWG or OpenUI.
  • Educate developers on CSS best practices, which is about implementation rather than the features themselves.
  • Manage browser compatibility data, which is already handled by Baseline.

That said, these exclusions don’t invalidate related efforts. For example, a CSS era could list all features specced in that period, with Baseline references for what shipped — feeding ideas into future Interop rounds. But the core focus for CSS-Next remains narrow: research how the community understands modern CSS, build a shared understanding of feature evolution since CSS3, group those features into digestible levels like CSS4 and CSS5, and educate the community about what’s current.

Where The Lines Are Drawn: CSS4, CSS5, CSS6

The initial plan was to pick up after CSS3 and label the next batch of features CSS4. The problem? There are far too many features to squeeze into a single level, even if you only count what shipped after 2020. So the working definition splits that gap, calling CSS4 a closed era ending around 2018 — features that are already fundamental to CSS today — and moving into the next logical period, CSS5.

Here’s how the eras are currently sketched out:

CSS3 (~2009-2012): The Level 3 specs as defined by the CSSWG. This era is immutable.

CSS4 (~2013-2018): Essential features that weren’t part of CSS3 but are now a fundamental part of the language.

CSS5 (~2019-2024): Newer features whose adoption is steadily growing.

CSS6 (~2025+): Early-stage features planned for future CSS.

We Need Your Input

Back in May, the group published a request for comments. So far, only a few responses have come in, and they’ve been taken into account — but a much broader response is needed to make the approach solid. The group is asking for community-wide feedback on whether these era boundaries make sense, where the lines should be drawn, and whether the stated goals are the right ones.

Anyone can join the CSS-Next group. There’s also an incubation group that meets biweekly on Mondays at 8:00 a.m. Pacific Time (2:00 p.m. GMT) for an hour-long session. The mission can seem abstract at first, but once you see the pace of change in CSS, the value becomes clear. Even seasoned developers are starting to admit they wouldn’t know where to begin with vanilla CSS on a fresh project — there’s no shame in that. Framing features by era may be a marketing exercise, but it’s one that gives everyone a better way to talk about where CSS has been and where it’s heading.

Resources

Smashing Editorial