Skip to main content
All WCAG success criteria

1.2.2 Captions (Prerecorded)

  • Level A
  • Perceivable
  • Since WCAG 2.0

Summary

This criterion protects people who are deaf or hard of hearing, and it also serves anyone watching in a noisy office, on muted public transport, or in a second language. It applies to synchronised media: prerecorded video that has an audio track. Every such video needs captions, whether closed (a track the viewer switches on) or open (text burned permanently into the picture). Both forms are acceptable. The single exception is a video that exists purely as an alternative to text already on the page and is clearly labelled as such.

The auditor's mental model is equivalence, not mere presence. Captions are not the same thing as subtitles: subtitles carry dialogue for people who can hear the soundtrack, while captions must also identify who is speaking when it is not obvious and describe sounds that carry meaning (a phone ringing, applause, a warning tone). A caption track exists or it does not, but its quality is gradeable too. Raw auto-generated captions that garble names, mangle technical vocabulary, or drop punctuation until the meaning is lost do not meet the equivalence bar, and it is honest auditing to fail them: the criterion requires captions for the audio content, not a rough statistical guess at it.

Official wording

Captions are provided for all prerecorded audio content in synchronized media, except when the media is a media alternative for text and is clearly labeled as such.

Success Criterion 1.2.2 Captions (Prerecorded), Web Content Accessibility Guidelines (WCAG) 2.2, W3C Recommendation, 5 October 2023 (updated 12 December 2024). Copyright © 2023-2024 World Wide Web Consortium. https://www.w3.org/copyright/document-license-2023/. Reproduced unmodified under the W3C Document License.

EN 301 549 mapping

Web pages
Clause 9.1.2.2
Software and native apps
Clause 11.1.2.2

Clauses 9.1.2.2 (web) and 11.1.2.2 (software) apply this criterion unchanged, so the same requirement binds web pages and native app user interfaces.

Clause references are to EN 301 549 V3.2.1 (2021-03), the harmonised European standard. Descriptions are our own summary, not the text of the standard.

In practice

Web

Inventory the prerecorded videos on each page, including hero background videos with narration and videos inside embedded third-party players. For each one, first establish whether a caption mechanism exists: a CC control in the player, a track element with kind="captions" on a native video element (inspect in browser devtools), or open captions visible in the picture. A video whose only audio is a music bed with no speech and no meaningful sounds has no audio content to caption; note it rather than failing it.

Then sample for quality. Turn captions on and watch two or three passages, including any with names, numbers, or technical terms. Check accuracy against the soundtrack, check synchronisation (captions should track the speech, not lag several seconds behind), and check non-speech information: speaker identification in multi-voice content and meaningful sound effects. Uncorrected auto-captions usually reveal themselves quickly through garbled proper nouns and missing punctuation; grade what you observe, not the vendor's claim.

Finally, test the exception properly before applying it. The video must be an alternative for text that is present and complete on the page, and it must be clearly labelled as such. A promotional video that loosely overlaps with nearby copy is not a media alternative for text.

iOS

The European standard applies this criterion unchanged to software, so native apps are graded to the same bar as web pages.

AVPlayer supports caption tracks embedded in the stream or delivered as sidecar files (WebVTT and broadcast caption formats), and it exposes them through its media selection options. The key platform behaviour to verify is respect for the system preference: Settings, Accessibility, Subtitles and Captioning, Closed Captions + SDH. With that setting enabled, play every prerecorded video in the app and confirm captions appear automatically and honour the user's caption styling. Standard AVPlayer UI does this when a caption track is present; custom players must implement it, and often do not.

Failure spots on iOS: onboarding and marketing videos shipped without any caption track, custom players that hide or omit a caption toggle even when the media contains one, and players that ignore the Closed Captions + SDH setting so captions never appear unless the user hunts for an in-app switch. Use Accessibility Inspector or the player UI to confirm which tracks the media actually carries rather than trusting the design spec.

Android

The same European-standard position applies: this criterion carries over to Android apps unchanged.

ExoPlayer (now Media3) selects and renders caption tracks and, by default, honours the system caption settings exposed through the platform captioning service (Settings, Accessibility, Caption preferences on most devices). Enable system captions, set a distinctive style (large text, coloured background), then play each prerecorded video in the app. Confirm captions appear, remain in sync, and reflect the chosen style; a custom renderer that hardcodes its own caption appearance is ignoring the captioning service and worth reporting alongside any missing tracks.

Where the app uses a fully custom player, check whether caption tracks exist in the media at all and whether the UI offers any way to enable them. Grade the same quality points as on web: accuracy on a sample, speaker identification where voices change, and meaningful sound effects. Failure spots on Android mirror iOS: help videos without tracks, players with no caption control, and auto-generated tracks that were never reviewed.

Pass and fail examples

Passes:

  • A product demo video with a closed caption track that renders dialogue accurately, names each speaker in a two-person conversation, and includes cues like a notification chime that the presenter reacts to.
  • A short social clip with open captions burned into the picture: accurate, legible, and synchronised. Open captions satisfy the criterion just as closed captions do.
  • Auto-generated captions that were reviewed and corrected by a human, verified accurate on a sample including technical terms.
  • A video walkthrough that repeats the instructions written in full on the same page, clearly labelled as a video alternative to the text: the exception applies.

Fails:

  • A prerecorded webinar recording with no caption track and no open captions.
  • Uncorrected auto-captions that render the product name and key figures as gibberish: a track exists, but it is not an equivalent for the audio content.
  • Captions that carry the dialogue but never identify speakers in a panel discussion and omit meaningful sounds, leaving caption users to guess who said what.
  • A caption track that drifts several seconds out of sync by the middle of the video.
  • A native app video player that offers no way to enable captions and ignores the system caption preference, even though the media carries a track.

Not a fail under this criterion:

Commonly confused with

  • SC 1.2.1 Audio-only and Video-only (Prerecorded). A transcript satisfies 1.2.1 for audio-only and video-only media, but it does not satisfy this criterion for video with audio. Synchronised media needs captions that play in time with the picture; a transcript link below the player is good practice, not a substitute.
  • SC 1.2.4 Captions (Live). Same requirement, different timeline. Live broadcasts are graded under 1.2.4 (Level AA); anything published as a recording is graded here. Misfiling a video-on-demand finding under the live criterion understates the conformance level at stake, since this criterion is Level A.
  • SC 1.2.5 Audio Description (Prerecorded). Captions move the audio into text for people who cannot hear; audio description moves the visuals into speech for people who cannot see. A single video can fail both, and the evidence is different: check the caption track for one, the soundtrack's coverage of visual content for the other.
  • Subtitles versus captions. Translation subtitles assume the viewer can hear the soundtrack, so they typically omit speaker identification and sound effects. A subtitle track alone does not automatically meet the captions bar; grade it against the full audio content, not just the dialogue.

How AUDITSU tests this

AUDITSU's audit walkthrough surfaces media checks on every screen that contains video. For each prerecorded video you record whether a caption mechanism exists (closed track, open captions, or a labelled media alternative for text), then the walkthrough prompts the quality sample: accuracy on a spoken passage, synchronisation, speaker identification, and meaningful sound effects. On native app screens it also prompts the platform check, playing the video with the system caption preference enabled. Each check records a pass, fail, or not applicable result per screen with your evidence attached, so a garbled auto-caption track lands in the report as a graded finding against the exact video, not a vague note.

Because the European standard applies this criterion unchanged to both web content and software, the same walkthrough questions serve web, iOS, and Android audits without remapping.

For the full guided workflow, see the audit platform.

This page explains a standard requirement and how we test it in practice. It is guidance, not legal advice. For a formal conformance assessment, consult a qualified accessibility auditor.

WCAG 2.2: W3C Recommendation, 5 October 2023 (updated 12 December 2024).