WCAG 2.2 vs WCAG 3.0 in 2026 | Which Standard Should You Actually Build To?
Last updated: 2026-07-29
Sooner or later someone in a meeting says 'should we wait for WCAG 3.0?' and the room pauses, because nobody wants to spend a budget cycle conforming to a standard that is about to be replaced. The short answer is no, and understanding why saves you from a genuinely costly mistake. WCAG 2.2 is a finished W3C Recommendation, published in October 2023 - it is done, it is stable, and it is what accessibility law and procurement actually reference today. WCAG 3.0, properly called the W3C Accessibility Guidelines 3.0, is a Working Draft: an in-progress document that has been iterated on publicly for years, whose conformance model has already been substantially rethought more than once, and which the W3C itself has been clear is not expected to become a Recommendation for a long time yet. Nothing anywhere requires it. The two are not even the same kind of document: 2.x is a checklist of testable success criteria organized under four principles, each pass or fail; 3.0 proposes a different structure built around outcomes and a more nuanced way of expressing how well a site does, intended to cover things binary pass/fail testing handles badly - like whether a process is genuinely usable end to end, not just whether every element has a label. That is a worthwhile direction, and it is also years of standards work away from being something a regulator can point at. This comparison sets out what each document is, what the law says today, and what to do in the meantime. None of this is legal advice; consult a qualified attorney for your jurisdiction.
At a Glance
| Feature | WCAG 2.2 | WCAG 3.0 (W3C Accessibility Guidelines 3.0) |
|---|---|---|
| Status | Finished W3C Recommendation (October 2023) | Working Draft - in progress, subject to change |
| Required by law anywhere | Yes - referenced directly or via EN 301 549 and national rules | No, and cannot be until it is finished |
| Structure | Four principles, testable success criteria, Levels A/AA/AAA | Proposed outcomes and assertions with a non-binary assessment model |
| Scope | Web content | Proposed to extend beyond web pages to apps and other digital products |
| Cognitive and learning disability coverage | Improved in 2.2 but still the weakest area | An explicit priority of the draft |
| Backwards compatibility | Meeting 2.2 AA also means meeting 2.1 AA and 2.0 AA | A different model; not a drop-in upgrade path |
| Tooling, audits, and VPATs available | Mature ecosystem across every vendor | None built around it yet |
| Expected timeline | Stable now; use it today | Years from Recommendation, plus a long adoption period after that |
| What to do about it in 2026 | Build and audit against 2.2 Level AA | Read it for direction; do not plan against it |
WCAG 2.2
Pros
- It is finished and stable - a Recommendation, not a draft - so what you build to today will not be invalidated by an editor's revision next quarter
- It is backwards compatible with WCAG 2.1 and 2.0: meeting 2.2 Level AA means you also meet 2.1 Level AA and 2.0 Level AA, so work done against those earlier versions is not wasted
- This is what the law and procurement reference in practice - EU obligations flow through the harmonised standard EN 301 549, the US Department of Justice's 2024 ADA Title II rule sets WCAG 2.1 Level AA for state and local government, and settlements routinely name a WCAG 2.x Level AA target
- Every testing tool, audit template, VPAT form, training course, and consultant on the market is built around 2.x success criteria, so you can buy help and compare quotes
- The nine criteria added in 2.2 target very real modern failures - focus hidden behind sticky headers (2.4.11), drag-only interactions (2.5.7), tap targets too small to hit (2.5.8), help that moves around between pages (3.2.6), forms that make you retype what you already entered (3.3.7), and logins that demand you solve a puzzle or recall a password from memory (3.3.8)
Cons
- Pass/fail success criteria measure conformance, not experience - a site can technically satisfy every criterion and still be exhausting to use with a screen reader
- Coverage of cognitive and learning disabilities is thinner than for vision and motor disabilities, which is one of the gaps 3.0 is explicitly trying to address
- Some criteria require judgement rather than measurement, so two competent auditors can disagree on the same page
- Fully conforming at Level AA is genuinely demanding for a large site, and the all-or-nothing conformance model gives no credit for substantial partial progress
WCAG 3.0 (W3C Accessibility Guidelines 3.0)
Pros
- Designed to cover more than just web pages, extending toward apps, tools, and other digital products in one coherent model
- Moves away from strict binary pass/fail toward assessing outcomes, which is a better fit for questions like 'can a person actually complete this checkout' rather than 'does every input have a label'
- Aims to give substantially better coverage of cognitive, learning, and neurological disabilities, the area where 2.x is weakest
- Being developed openly, so organizations can read the drafts, follow the reasoning, and comment on proposals as they form
- Reading the current draft is a genuinely useful signal of where accessibility practice is heading, even though none of it is binding
Cons
- It is a draft, and drafts change - the conformance and scoring model in particular has been reworked more than once, so anything you build against it today may not match the final text
- No law, regulation, procurement standard, or settlement anywhere requires WCAG 3.0, and none can until it is a finished Recommendation
- The W3C has been consistent that it is not expected to reach Recommendation status for years, and the transition period after that will be longer still
- No audit tooling, VPAT template, or established consultancy methodology is built around it, so there is nothing to buy and no one to hire against it
- Treating it as a reason to delay is the single most expensive misreading of it - the risk is entirely present-tense, and it is measured against 2.x
Our Verdict
Build to WCAG 2.2 Level AA and stop thinking about 3.0 as a deadline. WCAG 3.0 is a Working Draft whose own conformance model is still being redesigned, it is not referenced by any law or contract anywhere, and the W3C has been clear that a finished Recommendation is years out - after which regulators would still need years to adopt it. Meanwhile your actual exposure is present-tense and measured against 2.x: the European Accessibility Act's obligations have been in force since June 2025, the US Department of Justice's Title II rule sets a WCAG 2.1 Level AA target for state and local government with compliance deadlines already running, and private ADA settlements routinely specify WCAG 2.x Level AA. There is also a quietly reassuring detail in how the 2.x line works: every version is backwards compatible, so a site conforming to 2.2 Level AA also conforms to 2.1 and 2.0 at that level. The work is not throwaway. And the direction 3.0 points - can a real person actually complete the task, not merely does each element have a label - is something you can adopt right now without waiting for anyone to publish anything, by watching someone use your site with a screen reader and fixing what stops them. Start with the nine criteria 2.2 added; they describe failures that are on your site today. None of this is legal advice; consult a qualified attorney for your jurisdiction.
Further Reading
Other Comparisons
Get our free accessibility toolkit
We're building a simple accessibility checker for non-developers. Join the waitlist for early access and a free EAA compliance checklist.
No spam. Unsubscribe anytime.