When creating accessible web content, you need to test your work — and while automated tools can find many of the barriers that affect users with disabilities, manual testing is essential.
The World Wide Web Consortium (W3C) recommends using both manual and automated tests to ensure conformance with the Web Content Accessibility Guidelines (WCAG). WCAG is widely considered the international standard for online accessibility.
In most cases, accessibility experts test content for WCAG conformance by using screen readers, software that converts text to audio or braille. But what screen reader is the best choice for testing content against WCAG?
The Most Popular Screen Readers for Accessibility Testing
When selecting screen readers for accessibility testing, it’s important to use tools that are popular with real-life users. WebAIM (Web Accessibility In Mind) performs an annual survey of screen reader users, which includes questions about software preferences and demographics.
In WebAIM’s 2021 survey, these were the most popular primary screen readers:
- JAWS (Job Access With Speech) - Used by 53.7% of survey respondents.
- NVDA (NonVisual Desktop Access) - Used by 30.7% of survey respondents.
- Apple VoiceOver - Used by 6.5% of survey respondents.
Since the goal of accessibility testing is to understand the experiences of real users, this data shows two clear choices for accessibility testing: JAWS and NVDA.
Can I Test Content for Accessibility with JAWS or NVDA?
Ideally, manual screen reader tests should be performed by experts who have disabilities. We’ve explained why this is important in other articles, but here’s the short version: People who live with disabilities understand how assistive technologies work. They can provide feedback with context, which can help developers address accessibility barriers more effectively.
With that said, downloading a screen reader can give you valuable perspective. NVDA can be especially useful, as it’s free, portable, and compatible with most Microsoft Windows applications. While JAWS is a more popular tool, it’s fairly expensive (at the time of publication, a home license for JAWS costs $95 per year).
We’ve written in-depth articles with tips for reviewing content with both screen readers:
- JAWS Screen Reader: An Overview for Developers and Content Creators
- NVDA Screen Reader: An Overview for Developers and Content Creators
Additionally, Apple VoiceOver and Android Talkback can be useful tools for testing certain types of content (especially if you’re creating a mobile app or mobile-specific website). Here are resources for understanding these tools:
- How Apple VoiceOver Is Used For Mobile Accessibility Testing
- Google TalkBack: An Overview of Android's Free Screen Reader
When using any of these assistive technologies, remember that your experience won’t necessarily align with the experiences of regular screen reader users. People who use screen readers every day understand how to navigate web content quickly. They know how the software works, and they can determine whether a certain issue is a serious problem or a minor inconvenience.
You won’t have the same aptitude. As a result, you might miss accessibility barriers. Alternatively, you might assume that your entire website is inaccessible when you actually need a few minor tweaks to follow WCAG.
Manual accessibility testing shouldn’t rely on a single screen reader
At the Bureau of Internet Accessibility (BOIA), our experts test content with several screen readers when performing manual evaluations. We believe this approach provides the most accurate results.
Here’s why: Different screen readers may vary in their support of WAI-ARIA (Web Accessibility Initiative - Accessible Rich Internet Applications), a set of attributes that define web elements for screen readers and other assistive tools. Additionally, screen reader output may vary depending on the software’s interpretation of semantic HTML.
By combining automated testing with manual tests performed by users with visual disabilities, we’re able to provide comprehensive feedback — along with detailed suggestions for remediation. Read more about our four-point hybrid testing methodology.