Dark Mode as a Compliance Method for Non-text Contrast

[youngheek]

I would like to inquire whether supporting a Dark Mode feature in a mobile environment to enhance contrast can be interpreted as a valid approach to comply with WCAG Success Criterion 1.4.11 Non-text Contrast.

Success Criterion 1.4.11 mandates contrast for all visual elements, including user interface components. However, strictly adhering to the minimum contrast requirements for every design element can excessively constrain visually complex designs or branding elements, potentially hindering design flexibility and aesthetic appeal.

Therefore, we request an official interpretation on whether providing users with an option to select a high-contrast theme, such as Dark Mode, can be accepted as an alternative or complementary method for fulfilling SC 1.4.11, rather than uniformly raising the contrast of all design elements to 3:1 or higher.

[GreggVan]

Well if you can guarantee that dark mode on every device would cause the non-text contrast to be met - perhaps.

But dark mode usually just means the background is dark - not that all content will become black and white -- or that content will have high contrast. Most often - dark mode will not result in content that has low contrast from having high contrast -- though it might.

Again - you would have to check it with all dark modes that might be used on different devices to view your content. AND be sure that all the other success criteria are still met in that mode.

I would not guess that you could rely on dark mode to solve the contrast for a page that did not meet contrast. (and in fact dark mode might cause some pages that meet contrast to fail it.)

So I would not think this would be a viable solution.

I am interested in other's thoughts on this

[youngheek]

Thank you for your answer. If a service provider offers a Dark Mode (e.g., using specific CSS colors for Dark Mode) that meets the required luminance contrast for accessibility, can this be considered sufficient for meeting the contrast compliance criteria?

[patrickhlauke]

the path to / controls to enable dark mode need to have sufficient contrast themselves (assuming we're talking about a switch on the actual page/app). and you can't just rely on "if the users have colour vision deficincies, they'll have dark mode enabled at OS level".

[TestPartners]

I don't think this is viable unless the dark mode styles are all specified by the author. If you leave anything to the user agent, you can't guarantee the colour contrast will be sufficient.

It's worth noting that dark mode isn't a high contrast theme - the contrast is often lower than the default theme. The purpose of dark mode is primarily to reduce eye strain, which is an entirely different issue.

[WayneEDick]

I cannot speak for all users with low vision, but dark mode can be very helpful with text because the quantity of light is much smaller than the quantity of dark. Thus you can get greater contrast with less overall light. But often illustrations create the reverse problem. So I swap polarity when I have to study an image. The ability to swap polarity is important. The need for high luminosity needed to create contrast ratio is important for all people with LV and light sensitivity. So the issue from a user's POV is flexibility. Best, Wayne

…

On Mon, Nov 17, 2025 at 10:03 PM Gregg Vanderheiden < ***@***.***> wrote: Well if you can guarantee that dark mode on every device would cause the non-text contrast to be met - perhaps. But dark mode usually just means the background is dark - not that all content will become black and white -- or that content will have high contrast. Most often - dark mode will not result in content that has low contrast from having high contrast -- though it might. Again - you would have to check it with all dark modes that might be used on different devices to view your content. AND be sure that all the other success criteria are still met in that mode. I would not guess that you could rely on dark mode to solve the contrast for a page that did not meet contrast. (and in fact dark mode might cause some pages that meet contrast to fail it.) So I would not think this would be a viable solution. I am interested in other's thoughts on this — Reply to this email directly, view it on GitHub <#418 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AB6Q4F3PJW4PUCMVHNUMFJD35KZDNAVCNFSM6AAAAACMNHEIZSVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTIOJZG4ZDINA> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>