Check a page with keyboard, zoom and a screen reader

Choose a real task to test

Open the page preview and decide what a visitor needs to do: read an event notice, open a report, move through a gallery or register. Check the final public version again after publication. These checks find common barriers; they are not a complete conformance audit.

Record the URL, date, browser, viewport and result. If a step fails, describe what happened and send it to the responsible editor or website team. Do not simply mark the entire page “inaccessible” without identifying the blocked task.

Put the mouse aside

1. Reload the page. Press Tab to move forward and Shift+Tab to move back. Every interactive control should be reachable in a sensible order.
2. Look for a visible focus indicator. Make sure cookie notices, sticky headers or floating controls do not completely hide the focused item.
3. Use Enter for links and Enter or Space for buttons. Menus and other complex controls may also use arrow keys.
4. Open and close any menu or gallery viewer. Escape should close a dialog where that behavior is supported; focus should return to a sensible place.
5. Complete the intended task without getting trapped. Test errors and confirmation for a form in an approved test environment.

Some browsers, especially on macOS, have keyboard-navigation settings that affect which controls receive Tab focus. Confirm those settings before interpreting skipped controls as a site defect.

Increase text size and narrow the page

Check browser zoom at 200% for clipped text, overlap and lost controls. For reflow, test a viewport 320 CSS pixels wide, or start with a 1280-CSS-pixel-wide viewport and zoom to 400%. Ordinary paragraphs should not require scrolling both horizontally and vertically. Content that genuinely needs two-dimensional layout, such as a data table, has specific exceptions.

Try the page on a small screen too. Confirm that links are comfortable to activate and that orientation is not unnecessarily restricted. A narrow-screen test is useful, but does not replace zoom and text-resize checks.

Review meaning as well as appearance

Read the headings as an outline. Check that link labels identify their purpose and that essential event information is outside the flyer image. Ask whether an image description supplies the information its context requires. For video, review captions for accuracy and determine whether important visual information needs audio description.

Check text and control contrast using a measurement tool. Normal text generally needs 4.5:1; large text has a 3:1 threshold. The example tools include a solid-color contrast check. Image backgrounds, hover states and gradients need additional inspection.

Add trained and automated testing

A trained reviewer should check headings, landmarks, links, form labels and status announcements with a screen reader, then complete the main task. Record the browser and screen-reader combination. Merely hearing the page read aloud does not establish that its structure and controls are accessible.

Run an automated scanner on the same page and interactive states. Confirm each finding, investigate items needing manual review, and retain the results. Automated tools cannot reliably judge the quality of every description, caption, reading order or interaction. Combine their output with manual testing and, where practical, feedback from people with disabilities.