Headings Screen reader compatibility

Last updated: September 5, 2016

Screen reader compatibility for HTML and ARIA headings, showing how failures and techniques work in specific screen reader / browser combinations.

The results include two types of test:

Works as expected

These tests use conformant HTML or WCAG sufficient techniques, and work in all tested browser / screen reader combinations.

Screen ReaderNVDAJAWSVoiceOverWin-EyesDolphinSaToGo
BrowserIEFFIEFFMaciOSIEIEIE
Should work. Works in 100% Heading is IMG with ALTGoodGoodGoodGoodGoodGoodGoodGoodGood

Sometimes works

These tests use conformant HTML or WCAG sufficient techniques and might be expected to work in screen readers. This doesn't always happen.

Screen ReaderNVDAJAWSVoiceOverWin-EyesDolphinSaToGo
BrowserIEFFIEFFMaciOSIEIEIE
Should work. Fails in 1% - 25% ARIA role=headingBetterGoodGoodGoodGoodGoodBetterBadBad

Expected to fail

These tests use non-conformant HTML or WCAG failures and are expected to fail in screen readers.

Screen ReaderNVDAJAWSVoiceOverWin-EyesDolphinSaToGo
BrowserIEFFIEFFMaciOSIEIEIE
Should fail. Fails in 1% - 25% Empty headingGoodGoodGoodGoodGoodBetterGoodGoodBad
Should fail. Fails in 76% - 100% Heading is IMG ALT=""BadBadBadBadBadBadBadBadBad
Should fail. Fails in 76% - 100% Heading is IMG ALT=filenameBadBadBadBadBadBadBadBadBad
Should fail. Fails in 76% - 100% Heading is IMG with no ALTBadBadBadBadBadBadBadBadBad
Should fail. Fails in 1% - 25% Heading non-breaking spacesGoodBetterGoodBetterBadGoodBetterGoodBad
Should fail. Fails in 76% - 100% Heading only punctuationBadBadBadBadBadBadWorseBadBad
Should fail. Fails in 1% - 25% Heading only spacesGoodGoodGoodGoodGoodBetterGoodGoodBad
Should fail. Fails in 76% - 100% Nested ARIA headingsBadBadBadBadBadBadBadBadBad
Should fail. Fails in 76% - 100% Nested headingsBadBadBadBadBadBadBadBadBad

Key

Tests expected to fail (due to authoring errors) are marked with Expected to Fail.

Test notes

The threshold for inclusion in these results is 5% usage in the most recent WebAIM screen reader survey. Chrome and Android still fall below the 5% threshold.

All tests were carried out with screen reader factory settings. JAWS in particular has a wide variety of settings controlling exactly what gets spoken.

Screen readers allow users to interact in different modes, and can produce very different results in each mode. The modes used in these tests are:

In the «What the user hears» column: