WCAG Requirements for Website Video

The success criteria that apply to video

WCAG is organized into Guidelines, such as Guideline 1.2 Time-based Media, which contain numbered Success Criteria. The criteria are the testable requirements. Guideline 1.2 contains the ones specific to video.

Two general criteria also apply to how video is delivered:

• 2.1.1 Keyboard (Level A). Every player control (play, pause, volume, captions, fullscreen, scrub) must be operable by keyboard alone.
• 2.3.1 Three Flashes or Below Threshold (Level A). No content flashes more than three times per second, unless it stays below the general and red flash thresholds. The Trace Center's Photosensitive Epilepsy Analysis Tool tests for this.

Captions and audio descriptions are different requirements

This is the most common point of confusion, and the structure of the criteria makes it easy to miss.

Captions carry the audio to someone who cannot hear it. A caption track that only transcribes dialogue is incomplete: SC 1.2.2 also expects speaker identification and notation of significant non-speech audio, such as a door closing or an alarm.

Audio descriptions carry the visuals to someone who cannot see them. If your video shows a chart, a screen recording, an on-screen phone number, or a demonstration where the narrator says "like this," none of that reaches a blind user without description.

The Level A and Level AA requirements differ in an important way. At Level A, SC 1.2.3 lets you satisfy the visual-information obligation with either an audio description or a complete text alternative that describes the video. At Level AA, SC 1.2.5 removes the text option and requires the audio description itself.

Since the DOJ's Title II rule and most settlement agreements specify Level AA, organizations targeting that standard need actual audio description, not a transcript.

The practical implication for production: describing visual action is far cheaper to plan than to retrofit. A script written so the narration already covers what is on screen can reduce or eliminate the need for a separate description track.

Automatic captions usually do not qualify

Platform auto-captioning from YouTube, Vimeo, Zoom, or similar services produces a starting point, not a compliant caption track. Typical failures include misrecognized proper nouns and technical terms, missing or wrong punctuation that changes meaning, no speaker identification, and no notation of non-speech audio.

Auto-generated captions need review and correction before they meet SC 1.2.2. Platform caption editors make this straightforward, and for a short business video it is usually a matter of minutes.

Three criteria that are often cited incorrectly

These appear on many video accessibility checklists, including ones published by accessibility vendors, applied in ways the criteria do not actually support.

1.4.3 Contrast (Minimum) covers text and images of text. Text burned into video frames generally falls under the criterion's exception for text that is "part of a picture that contains significant other visual content," so it is not measured the way body copy is. Captions rendered as text by the player are a different matter and are subject to the requirement. Where contrast does apply, the thresholds are 4.5:1 for normal text and 3:1 for large text, which WCAG defines as 18pt, or 14pt bold.

1.4.10 Reflow is not a general responsive design requirement. It requires content to be presentable at 320 CSS pixels wide without two-dimensional scrolling. Orientation is governed separately by SC 1.3.4, which requires content to work in both portrait and landscape unless a specific orientation is essential.

2.2.1 Timing Adjustable governs time limits the content imposes on the user, such as session timeouts and auto-advancing carousels. It is not a requirement to provide playback speed controls. Variable playback speed is a good feature, and it is not what this criterion asks for. Auto-starting video that moves for more than five seconds is covered by SC 2.2.2 Pause, Stop, Hide (Level A).

Testing

Automated tools catch a narrow slice of this. WAVE will flag a missing caption track and player markup problems; it cannot tell you whether captions are accurate, whether speakers are identified, or whether a description covers what matters on screen.

Practical checks that do not require specialist tooling:

1. Unplug the mouse. Tab to the player and operate every control. If you cannot reach the caption toggle, SC 2.1.1 fails.
2. Mute the video and watch it. If you cannot follow it, captions are doing their job or they are not.
3. Turn away from the screen and listen. Anything you miss is what audio description is for.
4. Read the caption file as text. Errors that are easy to gloss over on screen are obvious in a transcript.

How we handle video on client sites

We check video against these criteria as part of routine content review, and we flag videos that need caption correction before they go live. Where a client supplies video from a third-party producer, caption and description quality is worth specifying in the production contract rather than fixing afterward.

We also maintain platform-specific walkthroughs for YouTube and Vimeo, which cover where the caption controls live in each interface.

This is general guidance on a technical standard, not a legal opinion about any particular site. For broader context on how WCAG relates to ADA obligations, see our post on what ADA web accessibility actually requires.

Sources

• W3C, Web Content Accessibility Guidelines (WCAG) 2.1
• W3C Web Accessibility Initiative, Understanding WCAG 2.2, with per-criterion explanations including 1.2.3 Audio Description or Media Alternative, 1.2.5 Audio Description, and 1.4.3 Contrast (Minimum)
• W3C Web Accessibility Initiative, Making Audio and Video Media Accessible
• Trace Research and Development Center, Photosensitive Epilepsy Analysis Tool (PEAT)
• WebAIM, Captions, Transcripts, and Audio Descriptions
• Risingline, What ADA Web Accessibility Actually Requires