November 9, 2026
Video captions accessibility requirements: what WCAG 1.2.2 expects
The core video captions accessibility requirements come from WCAG Success Criterion 1.2.2, "Captions (Prerecorded)", a Level A criterion. It requires captions for all prerecorded audio content in synchronized media, except when the media is a media alternative for text and is clearly labeled as such. In practice, your videos must be playable with synchronized text that conveys what can be heard, so that people who are deaf or hard of hearing can understand the audio. That text goes beyond dialogue: it also identifies who is speaking and includes non-speech information conveyed through sound, such as meaningful sound effects or music. Below is what the W3C expects, when captions are not needed, the recognized techniques (including the track element) and the mistakes that cause a failure.
What WCAG 1.2.2 says about video captions
Success Criterion 1.2.2 applies to prerecorded audio content in synchronized media. The WCAG glossary defines synchronized media as audio or video synchronized with another format for presenting information and/or with time-based interactive components. Prerecorded simply means information that is not live. Live means information captured from a real-world event and transmitted to the receiver with no more than a broadcast delay; it falls under a separate criterion, 1.2.4 "Captions (Live)".
The W3C states that the intent is to enable people who are deaf or hard of hearing to watch synchronized media presentations. Captions provide the part of the content available via the audio track. The benefit is direct: through captions, these users can access the auditory information in the media.
The W3C also acknowledges that creating captions for time-sensitive material can currently be difficult. Authors may face a choice between delaying the information until captions are available, or publishing content that is inaccessible to deaf users, at least until captions are ready. It adds that, over time, captioning tools and building captioning into the delivery process can shorten or eliminate such delays.
What accessible captions must include
The WCAG glossary defines captions as a synchronized visual and/or text alternative for both speech and non-speech audio information needed to understand the media content. It explains that captions differ from dialogue-only subtitles: they convey the content of spoken dialogue, and also equivalents for non-dialogue audio information needed to understand the program.
The W3C examples make this concrete. In a video tutorial showing how to tie a knot, the captions read "(music)" followed by "Using rope to tie knots was an important skill for the likes of sailors, soldiers and woodsmen." For an orchestra's performance videos, captions render dialogue and lyrics verbatim, and identify non-vocal music by title, movement, composer and any information that helps users understand the nature of the audio, down to a description such as "Calm melody with a slow tempo".
- Spoken content (dialogue, narration, song lyrics).
- Identification of who is speaking.
- Meaningful sound effects.
- Music, laughter and speaker location when needed to understand the content.
- A display that, per the WCAG glossary, should not obscure or obstruct relevant information in the video.
Captions, closed captions and transcripts: the differences
WCAG distinguishes two forms. Closed captions are equivalents that can be turned on and off with some players. Open captions cannot be turned off, for example when they are images of text embedded in the video. Both are sufficient techniques for the criterion. The glossary also notes that in some countries captions are called subtitles. Our advice: check that you are dealing with complete captions and not dialogue-only subtitles.
What about a transcript? The deciding factor is synchronization: by the WCAG definition, captions are synchronized with the media. A text published next to the video without being displayed in time with the audio therefore does not match that definition. All the sufficient techniques listed for 1.2.2 involve providing captions, open or closed. The only exception is when the video presents no more information than text already on the page and is clearly labeled as an alternative to it, as explained below.
Finally, audio description, which is narration added to describe important visual details, can be captioned but does not need to be, since it describes information already presented visually.
Recognized techniques, including the track element
The W3C lists the techniques the working group deems sufficient to meet the criterion. A technique may go beyond the minimum requirement, and there may be other ways to meet the criterion. Two techniques are listed as sufficient: G93, providing open (always visible) captions, and G87, providing closed captions. G87 is implemented using any of the following techniques.
- H95: using the track element to provide captions.
- SM11: providing captions through synchronized text streams in SMIL 1.0.
- SM12: providing captions through synchronized text streams in SMIL 2.0.
- Using any readily available media format that has a video player supporting closed captioning.
When captions are not needed
Captions are not required when the synchronized media is itself an alternate presentation of information that is also presented as text on the web page. The W3C cites a video that presents no more information than the page text, but is easier for people with cognitive, language or learning disabilities to understand: since the information is already available in text or text alternatives, the video does not need captions.
Two examples illustrate this: a complex legal document where several paragraphs each come with a clip of a person speaking that paragraph's content, and an instruction manual describing a part and its orientation, accompanied by a clip showing the part in its correct orientation. In both cases, no captions are provided. Be careful, though: the media must be clearly labeled as an alternative to the text, otherwise the criterion fails.
Common failures of WCAG 1.2.2
The W3C working group lists three common failures. They make a good checklist before publishing a video. There is also a formal test rule for this aspect: "Video element auditory content has accessible alternative". For an audit, you can start with a free test of your pages, then set up continuous monitoring to catch new videos published without captions.
- F8: captions omitting some dialogue or important sound effects.
- F75: synchronized media without captions when it presents more information than the page.
- F74: synchronized media that is an alternative to text but is not labeled as an alternative.
Frequently asked questions
Are captions required for accessible videos?
Under WCAG 1.2.2, a Level A criterion, captions must be provided for all prerecorded audio content in synchronized media. The only exception is media that is an alternative for text and clearly labeled as such.
What should accessible captions include?
They convey dialogue, identify who is speaking and include the non-speech audio information needed to understand the content: meaningful sound effects, music, laughter, speaker location.
What is the difference between captions and a transcript?
Under WCAG, captions are synchronized with the media. A text published next to the video without synchronization does not match that definition, and the sufficient techniques for 1.2.2 all involve open or closed captions.
Are burned-in (open) captions acceptable?
Yes. Open captions, which cannot be turned off, are a sufficient technique (G93), just like closed captions (G87). In every case, the WCAG glossary notes that captions should not obscure relevant information in the video.
How do I add captions to an HTML video?
Sufficient technique G87 (closed captions) can be implemented with H95, which uses the track element to provide captions. You can also use a readily available media format whose player supports closed captioning.
Does WCAG 1.2.2 apply to live video?
No. 1.2.2 covers prerecorded content, meaning content that is not live. Live captions fall under Success Criterion 1.2.4.
Further reading on Conformly
Sources
This article relies on the official sources listed at the end of the page.