What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
An accessibility scanner finding is a lead to investigate, not proof of a violation—and a clean scan is not proof that a site is accessible. Confirm the relevant WCAG criterion, reproduce the interface state, and review the element in context before dismissing a finding. Then use manual and assistive-technology testing to look for problems automation can miss.
What counts as an accessibility testing false positive?
The UK Department for Education defines false positives as issues flagged by testing tools that are not actually issues when reviewed manually. That determination depends on the page, the element’s purpose, the applicable criterion, and the user experience—not simply on whether a developer can suppress the warning. The Department for Education’s guidance gives examples involving image meaning, link context, and contrast exemptions.
As an Amazon Associate I earn from qualifying purchases.
Keep false positives distinct from false assurance. A false positive is a reported issue that review finds is not an issue. False assurance occurs when a check passes—or a report is empty—even though a meaningful accessibility problem remains. A useful evaluation must guard against both.
Recommended Free Tools
Why do accessibility checkers report false positives?
They cannot reliably infer purpose and context
A rule can identify that an image has alternative text, but it cannot always decide whether that text conveys the information the image contributes. Likewise, a checker may assess link text without understanding whether the link’s purpose is clear in its surrounding content. Review the actual content and user task before deciding whether a warning is mistaken.
#1 Best Overall
A standard may have an applicable exception
A contrast checker may flag a logo even when the cited contrast criterion exempts logos or branding. Verify the exact criterion and its exceptions for the case at hand; do not treat every reported contrast issue as either automatically valid or automatically exempt.
The scan may evaluate the wrong interface state
Automated tools inspect rendered content. A scan performed before a menu, dialog, or other interactive region is opened may not cover its contents. The axe-core API documentation advises making inactive or non-rendered regions visible before analysis. A result therefore applies to the state scanned, not every state the page can enter.
Tool configuration may not match the evaluation
Rules, exclusions, supported file types, and standards alignment affect what a tool reports. For Section 508 evaluations, Section508.gov’s overview of testing methods advises examining how a vendor defines and measures rules against the standards and expectations in use, including ruleset version control, customization, exclusions, severity, context, file fidelity, and workflow integration.
Why can a scan pass while an accessibility problem remains?
A presence check can confirm that an alt attribute exists without judging whether its value helps anyone. The Department for Education notes that classroom art labeled alt="car" or alt="image123" can pass an automated presence check while failing to provide useful context. Conversely, a decorative image can appropriately have an empty alternative, alt="". The meaningful question is whether the text alternative fits the image’s purpose in that context, not merely whether the attribute is populated.
The Department for Education says accessibility tools identify around 30% to 40% of issues. That figure describes the share of issues identified, not a false-positive rate, and it is not a benchmark for a particular tool. The authoritative sources reviewed do not establish comparable tool-by-tool false-positive rates under a shared test corpus and method, so a definitive scanner ranking by false-positive rate is not supported.
Section508.gov describes the underlying trade-off: automated tools cannot apply human subjectivity, so they may produce excessive false positives or, if configured to suppress those, check only a small portion of requirements. Detection coverage and false-positive rate are different measures. A blank report is not a sound goal if it is achieved by hiding findings or narrowing checks until important requirements go untested.
Rank #3
How to investigate and classify a finding
- Record what the scan actually tested. Keep the rule identifier, affected element, page or user flow, scan configuration and version, and the interface state at the time. This practical record supports a scoped evaluation and report; it is not a verbatim checklist from WCAG-EM.
- Reproduce the same state. Return to the page and repeat the interaction. If the finding concerns a menu or dialog, expose it and rerun the scan while it is visible.
- Read the rule and relevant criterion. Check whether the criterion applies to this element and purpose, including any genuine exception such as the logo contrast case. Avoid deciding from the rule name alone.
- Review both the code signal and the user outcome. Check whether image alternatives accurately convey relevant information and whether link purpose is understandable in context. Attribute presence alone does not establish that an alternative is appropriate.
- Choose and document a disposition. Mark a finding as a false positive only when review supports that conclusion. Otherwise fix it or record it as needing further evaluation. If recurring false alarms justify a configuration change, manage that change centrally and track exclusions so repeated results are not silently dropped.
- Check for issues the scan may not reveal. Conduct manual and assistive-technology checks as separate parts of the evaluation, including key user flows and interactive states.
- Report scope and limitations. State what was evaluated, which samples and methods were used, what was found, and what remains untested. A clean scan should not be presented as proof of total conformance.
How to reduce false alarms without hiding real problems
- Align rules with the evaluation. Check which standards and criteria the ruleset covers, which version is in use, and how the tool handles supported content types.
- Make exclusions visible and reviewable. Record why a rule or element is excluded, who owns the decision, and when it will be reviewed. Treat repeated warnings as evidence to investigate, not as a reason to suppress them automatically.
- Scan representative states. Include authenticated content where relevant and expose important menus, dialogs, and other dynamic regions before analyzing them. A tool’s results cannot cover content it never evaluates.
- Use findings with useful context. Prefer reports that identify the affected element and provide remediation guidance and severity information, rather than relying on a bare pass/fail result.
- Integrate repeatable checks into development. Automated checks can help catch regressions in acceptance tests and CI, but keep the ruleset versioned and pair automation with review of user-facing behavior.
- Balance precision with coverage. Assess whether a proposed configuration change reduces noise while preserving checks for meaningful requirements. Fewer findings alone do not show that the evaluation improved.
These are evaluation considerations, not a tested ranking of vendors. Section508.gov’s selection criteria include ruleset alignment, source-format fidelity, customization, version control, exclusions, severity, context, and workflow integration. For dynamic states, axe-core’s documentation adds the need to expose non-rendered content before analysis.
Use a mixed evaluation, not a scanner verdict
W3C’s WCAG Evaluation Methodology (WCAG-EM) 2.0 describes a technology-agnostic process: define scope, explore the product, select representative samples, evaluate those samples, and report results. It is a W3C Group Note with informative guidance—not a new normative requirement or a replacement for WCAG.
The UK Department for Work and Pensions’ testing guidance treats automated tools as useful early checks and also describes manual and assistive-software testing. Its process-specific direction should not be generalized into a universal legal requirement, but it illustrates why scanning is only one part of an evaluation.
For a defensible assessment, plan for repeatable automated checks alongside manual review and testing with assistive technology. Explain the scope and remaining gaps rather than implying that the scanner alone established conformance. W3C also discusses broader challenges with accessibility guidelines conformance and testing.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Or skip the browser setup
If you need a rendered page capture to document a finding or compare an interface state, ScreenshotNeo can return a screenshot or PDF from one GET request. Its consent cleanup accepts cookie banners like a visitor and removes more than 60 known consent platforms, newsletter popups, and chat widgets; those steps can be turned off. Bot checks, blank pages, timeouts, failed loads, and cache hits are not billed, and response headers say which page verdict and billing outcome applied. Its MCP server offers take_screenshot, get_page_info, and capture_pdf for AI agents and MCP clients.
For example, using cURL:
curl -G "https://api.screenshotneo.com/v1/shot" -d access_key=YOUR_API_KEY --data-urlencode url=https://example.com -o shot.webp
See the ScreenshotNeo documentation for request options. It offers 1,000 screenshots a month free with no card; paid plans start at $5 for 3,000. A screenshot can help document what was rendered, but it does not replace accessibility testing or establish conformance. Sign up for 1,000 free screenshots a month, with no card required.
Frequently Asked Questions
Can an automated accessibility scan prove my site is accessible?
No. A scan only evaluates the checks and rendered states it covers; use it alongside manual and assistive-technology evaluation, and report the scope and remaining gaps.
Is there a reliable ranking of accessibility scanners by false-positive rate?
The cited authoritative sources do not provide comparable tool-by-tool rates using a shared corpus and method, so they do not support a definitive ranking.
Free tools Windows power users keep installed
One-click scans. No signup required.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
