Accessibility fixes are cheapest just before publishing and most expensive just after. Once a course is live, every problem becomes a support request, a republish and possibly a learner who could not finish. A short, ordered review at the end of development catches most issues while they are still a quick edit. The difficulty is knowing what belongs in that review, and in what order, so nothing important depends on memory. What should an e-learning accessibility checklist cover before you publish a course?
A pre-publish e-learning accessibility checklist should cover text alternatives, color and contrast, structure and reading order, keyboard operation and focus, interactions and quizzes, timing, audio and video, motion, and text resizing, followed by testing with assistive technology and, ideally, with people with disabilities. Automated tools catch missing alt text, contrast failures and missing captions quickly. Humans must judge whether alt text is useful, focus order makes sense, captions are accurate and interactions work with a screen reader. Run the automated checks first and the manual ones after. Responsibility for the result stays with the team that publishes the course.
What Does an Accessibility Check Before Publishing Cover?
It covers whether every learner can perceive, operate and understand the course. The reference point for most teams is WCAG, the Web Content Accessibility Guidelines. The W3C's WCAG overview explains that WCAG 2 exists in versions 2.0, 2.1 and 2.2, each with testable success criteria at three levels: A, AA and AAA. Many organizations aim for level AA.

For a pre-publish pass, it helps to group those criteria into areas a course developer can check slide by slide:
- Text alternatives for images and other non-text content.
- Color and contrast.
- Structure: headings, reading order, language, meaningful button names.
- Keyboard operation and visible focus.
- Interactions, quizzes and feedback.
- Timing and timers.
- Audio and video: captions, transcripts, audio description.
- Motion, animation and flashing.
- Text resizing and zoom.
Accessibility is one half of the release check; the other is functional. Run this list alongside the steps you use to test the SCORM package, so both sign-offs happen before the LMS administrator sees anything.
A note on obligations: which standard and level your organization must meet depends on law, contracts and policy, and those vary by country and sector. Your legal counsel or the relevant official source can confirm what applies. This checklist describes good practice against WCAG, not legal sufficiency.
Which Checks Can Be Automated and Which Need a Human?
Machines are good at presence and measurement; people are needed for meaning. A useful way to plan accessibility testing for e-learning is to split each check by who can do it:
| Check | Automated tool | Human judgement |
|---|---|---|
| Alt text | Finds images with none | Decides whether it is useful or should be decorative |
| Contrast | Measures text against background | Checks text over photos and in every state |
| Captions | Finds media without captions | Checks accuracy, sync and speaker labels |
| Focus order | Lists the order | Decides whether the order makes sense |
| Keyboard | Limited | Tabs through and operates every control |
| Screen reader experience | Very limited | Listens to the course and follows it |
Run automated checks first. They are fast, and fixing their findings removes noise before the slower manual review. Then do the manual checks in the order below.
Tooling can take the measurable checks off your plate and enforce some of them. AuthorLearn Studio runs a pre-flight check that blocks publishing when text fails WCAG AA contrast, unless someone overrides it with a written reason, and it can run an automated accessibility audit of the published course. Those checks support your review; they do not replace it, and your team remains responsible for checking the course it publishes.
How Do You Check Text, Images and Color?
Start with what the learner reads and sees. Work through every slide, including layers, feedback panels and the results slide:
- Every informative image has alt text that conveys its purpose in context.
- Decorative images are marked decorative, so screen readers skip them.
- Any text inside an image also appears in its alt text or on the slide.
- Normal text meets a contrast ratio of 4.5:1, and large text (18 point, or 14 point bold) meets 3:1, in every button state.
- Color is never the only signal. Correct and incorrect feedback uses words or icons as well as green and red.
- Charts and diagrams have a short alternative plus a fuller description or data table.
- Slide titles and headings are marked as headings, so screen reader users can move by them.
- Reading order on each slide matches the visual order.
- The course language is set, so screen readers pronounce text correctly.
- Buttons and links have names that make sense out of context ("Next slide", not "Click here").
Then check resizing. WCAG asks that text can be resized to 200% without loss of content or function. Zoom the browser and confirm that text is still readable and nothing important is cut off.
How Do You Check Navigation and Interactions?
With the mouse put away. Use Tab, Shift+Tab, Enter, Space, the arrow keys and Escape to move through the whole course, then check:
- Every control can be reached and operated, including the player menu, resources and volume.
- Focus is always visible.
- Focus order follows the slide's meaning, with content before navigation.
- Focus never gets trapped.
- Opening a layer moves focus into it; closing it returns focus to the button that opened it.
- Every question type in the quiz can be answered and submitted from the keyboard.
- Drag-and-drop and hotspot activities have a keyboard method that tests the same knowledge.
- Feedback is announced or placed where a screen reader user will find it.
- Any timer can be turned off, adjusted or extended, unless the time limit is genuinely essential.
Timed quizzes deserve special attention. A quiz timer exists for a reason, but if speed is not what the course is assessing, a fixed limit mostly penalizes learners who use assistive technology or need more reading time.
How Do You Check Audio, Video and Motion?
By watching, listening and then doing each again without one of the senses. Check:
- All prerecorded video and narrated sequences have accurate, synchronized captions that identify speakers and important sounds.
- Audio-only content has a transcript.
- Important visual information in video is also available as audio description or text.
- Audio that plays automatically for more than a few seconds can be paused or muted.
- Moving, scrolling or auto-advancing content can be paused.
- Nothing flashes more than three times in any one second.
- Decorative animation can be reduced or switched off for learners who set a reduced motion preference.
Watch each media item once with the sound off, then listen once without looking. If either pass leaves you missing information, so will some of your learners.
Who Should Test the Course Besides You?
At least one other person, and ideally people who use assistive technology every day. The developer who built a course knows where everything is, which makes them the worst judge of whether it is easy to find.
A workable order of testers:
- A colleague doing a first review with the W3C's Easy Checks, which cover basics such as alt text, headings, contrast, visible keyboard focus, zoom and captions. The W3C is clear that these are a first review, not a full evaluation.
- A developer or tester using a screen reader. Windows includes Narrator, and free screen readers are available, so cost is not a barrier to a basic listen-through.
- An accessibility specialist for high-stakes or widely used courses.
- People with disabilities, whose experience often reveals problems no checklist predicts, such as instructions that make sense visually but not when heard.
A basic screen reader listen-through does not need expertise, just patience. Turn the screen reader on, close your eyes or turn away from the screen, and try to complete one lesson. Listen for images announced with unhelpful names, buttons announced only as "button", content read in a confusing order, feedback that never gets announced, and places where you cannot tell what to do next. Ten minutes of this usually produces a more useful list than an hour of looking at the slides.
Record findings in one shared list with slide, issue, severity and owner, and retest each fix rather than assuming it worked.
Agree on what severity means before testing starts, so the list can be sorted quickly. One workable scheme: a blocker stops a learner from completing the course or a required part of it (an unreachable submit button, a video with no captions); a major issue makes a task much harder but still possible (a confusing focus order); a minor issue is a rough edge (a slightly vague alt text). Blockers get fixed before publishing, majors usually do, and minors go on a list for the next revision.
An E-Learning Accessibility Checklist Only Works If It Can Stop a Publish
An e-learning accessibility checklist is only as good as its authority. If a failed item can be waved through under deadline pressure, the checklist becomes paperwork. Agree in advance which failures block release, such as missing captions or a control that cannot be reached by keyboard, and who may approve an exception.
It also helps to keep the checklist in the same place as the course files and to record who ran it, when, and against which version. When a learner reports a problem months later, that record shows whether the issue slipped past the review or arrived with a later edit, and the answer changes what you fix in the process.
Order the checks, automate the measurable ones, keep the human ones, and give the list the power to say "not yet". A course that passes that gate is one far more learners can actually finish.
