Guias de acessibilidade para cego
Aprenda como identificar e corrigir problemas de acessibilidade comuns sinalizados pelo Axe Core — para que suas páginas sejam inclusivas e utilizáveis por todos. Consulte também o nosso Guias de validação HTML.
When you add an ARIA attribute to an HTML element, the browser exposes that information through the accessibility tree so assistive technologies like screen readers can interpret it. If the attribute name is invalid — whether due to a typo like aria-hiden instead of aria-hidden, or a fabricated attribute like aria-visible that doesn’t exist in the spec — the browser won’t recognize it. The attribute is effectively dead code, and the accessibility enhancement you intended never reaches the user.
This is classified as a critical issue because the consequences can be severe. For example, if you misspell aria-required as aria-requried on a form field, screen reader users won’t be informed that the field is mandatory. If you misspell aria-expanded on a disclosure widget, blind and deafblind users won’t know whether a section is open or closed. Keyboard-only users who rely on screen readers are also affected when interactive states and properties fail to communicate correctly.
Related WCAG Success Criteria
This rule maps to WCAG Success Criterion 4.1.2: Name, Role, Value (Level A), which requires that for all user interface components, the name, role, and states/properties can be programmatically determined. Invalid ARIA attributes fail to communicate states and properties to assistive technologies, directly violating this criterion. This applies across WCAG 2.0, 2.1, and 2.2, as well as EN 301 549 (guideline 9.4.1.2).
How to Fix It
-
Audit your ARIA attributes. Review every attribute in your markup that starts with
aria-and confirm it matches a valid attribute name from the WAI-ARIA specification. -
Check for typos. Common mistakes include
aria-labelled-by(correct:aria-labelledby),aria-hiden(correct:aria-hidden), andaria-discribedby(correct:aria-describedby). -
Remove invented attributes. Attributes like
aria-visible,aria-tooltip, oraria-icondo not exist in the WAI-ARIA spec and will have no effect. -
Use tooling. IDE extensions, linters (like
eslint-plugin-jsx-a11y), and the axe DevTools browser extension can catch invalid ARIA attribute names during development.
Common Valid ARIA Attributes
Here are some frequently used ARIA attributes for reference:
-
Widget attributes:
aria-checked,aria-disabled,aria-expanded,aria-hidden,aria-label,aria-pressed,aria-readonly,aria-required,aria-selected,aria-valuenow -
Live region attributes:
aria-live,aria-atomic,aria-relevant,aria-busy -
Relationship attributes:
aria-labelledby,aria-describedby,aria-controls,aria-owns,aria-flowto -
Drag-and-drop attributes:
aria-dropeffect,aria-grabbed
Examples
Incorrect: Misspelled ARIA Attribute
<button aria-expandd="false">Show details</button>
The attribute aria-expandd is not a valid ARIA attribute. Screen readers will not announce the expanded/collapsed state of this button.
Incorrect: Non-Existent ARIA Attribute
<div aria-visible="true">Important announcement</div>
The attribute aria-visible does not exist in the WAI-ARIA specification. It will be completely ignored by assistive technologies.
Incorrect: Typo in a Relationship Attribute
<input type="text" aria-discribedby="help-text">
<p id="help-text">Enter your full name as it appears on your ID.</p>
The attribute aria-discribedby is a misspelling of aria-describedby. The input will not be associated with the help text for screen reader users.
Correct: Properly Spelled ARIA Attributes
<button aria-expanded="false">Show details</button>
<div aria-hidden="true">Decorative content</div>
<input type="text" aria-describedby="help-text">
<p id="help-text">Enter your full name as it appears on your ID.</p>
Each of these examples uses a valid, correctly spelled ARIA attribute that browsers and assistive technologies will recognize and process as intended.
Why This Matters
The autocomplete attribute does more than enable browser autofill — it programmatically communicates the purpose of a form field to assistive technologies. This information is critical for several groups of users:
-
Screen reader users rely on the announced field purpose to understand what information is being requested. Without a valid
autocompletevalue, the screen reader may not convey this context clearly. - Users with cognitive disabilities benefit from browsers and assistive tools that can auto-populate fields or display familiar icons based on the field’s purpose, reducing the mental effort required to complete forms.
- Users with mobility impairments benefit from autofill functionality that minimizes the amount of manual input required.
- Users with low vision may use personalized stylesheets or browser extensions that adapt the presentation of form fields based on their declared purpose (e.g., showing a phone icon next to a telephone field).
This rule maps to WCAG 2.1 Success Criterion 1.3.5: Identify Input Purpose (Level AA), which requires that the purpose of input fields collecting user information can be programmatically determined. The autocomplete attribute is the standard mechanism for satisfying this requirement in HTML.
How the Rule Works
The axe rule autocomplete-valid checks that:
-
The
autocompleteattribute value is a valid token (or combination of tokens) as defined in the HTML specification for autofill. -
The value is appropriate for the type of form control it is applied to (e.g.,
emailis used on an email-type input, not on a checkbox). - The tokens are correctly ordered when multiple tokens are used (e.g., a section name followed by a hint token followed by the field name).
The rule flags fields where the autocomplete value is misspelled, uses a non-existent token, or is applied in an invalid way.
How to Fix It
- Identify all form fields that collect personal user information (name, email, address, phone number, etc.).
- Check if the data type matches one of the 53 input purposes defined in WCAG 2.1 Section 7.
-
Add the correct
autocompletevalue to each matching field. Make sure:- The value is spelled correctly.
- It is appropriate for the input type.
-
If using multiple tokens, they follow the correct order: optional section name (
section-*), optionalshippingorbilling, optionalhome,work,mobile,fax, orpager, and then the autofill field name.
-
Set
autocomplete="off"only when you have a legitimate reason to disable autofill — and note that this does not exempt you from the rule if the field still collects identifiable user data.
Common autocomplete Values
Here are some of the most frequently used values:
| Purpose |
autocomplete Value |
|---|---|
| Full name |
name |
| Given (first) name |
given-name |
| Family (last) name |
family-name |
| Email address |
email |
| Telephone number |
tel |
| Street address |
street-address |
| Postal code |
postal-code |
| Country |
country |
| Credit card number |
cc-number |
| Username |
username |
| New password |
new-password |
| Current password |
current-password |
Examples
Incorrect: Missing or Invalid autocomplete Values
<!-- Missing autocomplete attribute entirely -->
<label for="name">Full Name</label>
<input type="text" id="name" name="name">
<!-- Misspelled autocomplete value -->
<label for="email">Email</label>
<input type="email" id="email" name="email" autocomplete="emal">
<!-- Invalid autocomplete value -->
<label for="phone">Phone</label>
<input type="tel" id="phone" name="phone" autocomplete="phone-number">
In the examples above, the first field has no autocomplete attribute, the second has a typo (emal instead of email), and the third uses a non-standard value (phone-number instead of tel).
Correct: Valid autocomplete Values
<label for="name">Full Name</label>
<input type="text" id="name" name="name" autocomplete="name">
<label for="email">Email</label>
<input type="email" id="email" name="email" autocomplete="email">
<label for="phone">Phone</label>
<input type="tel" id="phone" name="phone" autocomplete="tel">
Correct: Using Multiple Tokens
When a form has separate shipping and billing sections, you can use additional tokens to distinguish them:
<fieldset>
<legend>Shipping Address</legend>
<label for="ship-street">Street Address</label>
<input type="text" id="ship-street" name="ship-street"
autocomplete="shipping street-address">
<label for="ship-zip">Postal Code</label>
<input type="text" id="ship-zip" name="ship-zip"
autocomplete="shipping postal-code">
</fieldset>
<fieldset>
<legend>Billing Address</legend>
<label for="bill-street">Street Address</label>
<input type="text" id="bill-street" name="bill-street"
autocomplete="billing street-address">
<label for="bill-zip">Postal Code</label>
<input type="text" id="bill-zip" name="bill-zip"
autocomplete="billing postal-code">
</fieldset>
Correct: Named Sections with section-*
You can use custom section names to group related fields when the same type of data appears multiple times:
<label for="home-tel">Home Phone</label>
<input type="tel" id="home-tel" name="home-tel"
autocomplete="section-home tel">
<label for="work-tel">Work Phone</label>
<input type="tel" id="work-tel" name="work-tel"
autocomplete="section-work tel">
By using valid, correctly applied autocomplete values, you ensure that assistive technologies can communicate the purpose of each field to users, browsers can offer reliable autofill, and your forms meet the requirements of WCAG 2.1 Success Criterion 1.3.5.
Some users need to adjust text spacing to make content readable. People with low vision may increase letter or word spacing to reduce visual crowding. People with cognitive disabilities, dyslexia, or attention deficit disorders often struggle to track lines of text that are tightly spaced — increasing line-height, letter-spacing, or word-spacing can make reading significantly easier.
When text-spacing properties are set inline with !important, they gain the highest specificity in the CSS cascade. This means user stylesheets, browser extensions, and assistive technology tools cannot override those values. The text becomes locked into a fixed spacing that may be difficult or impossible for some users to read.
This rule relates to WCAG 2.1 Success Criterion 1.4.12: Text Spacing (Level AA), which requires that no loss of content or functionality occurs when users adjust:
- Line height to at least 1.5 times the font size
- Spacing following paragraphs to at least 2 times the font size
- Letter spacing to at least 0.12 times the font size
- Word spacing to at least 0.16 times the font size
If inline !important declarations prevent these adjustments, the content fails this criterion.
How to Fix It
The fix is straightforward: do not use !important on inline style attributes for line-height, letter-spacing, or word-spacing. You have a few options:
-
Remove
!importantfrom the inline style declaration. Without!important, users can override the value with a custom stylesheet. - Move styles to an external or embedded stylesheet. This is generally the best approach because it separates content from presentation and gives users more control through the cascade.
-
If
!importantis truly necessary, apply it in a stylesheet rather than inline. Inline!importantstyles are virtually impossible for users to override, while stylesheet-level!importantcan still be overridden by user!importantrules.
Note that other inline style properties like font-size are not flagged by this rule — only the three text-spacing properties (line-height, letter-spacing, word-spacing) are checked.
Examples
Incorrect: Inline styles with !important
These examples fail because !important on inline text-spacing properties prevents user overrides.
<!-- line-height with !important — cannot be overridden -->
<p style="line-height: 1.5 !important;">
This text is locked to a specific line height.
</p>
<!-- letter-spacing with !important — cannot be overridden -->
<p style="letter-spacing: 2px !important;">
This text has fixed letter spacing.
</p>
<!-- word-spacing with !important — cannot be overridden -->
<p style="word-spacing: 4px !important;">
This text has fixed word spacing.
</p>
<!-- Mixed: word-spacing is fine, but letter-spacing has !important -->
<p style="word-spacing: 4px; letter-spacing: 2px !important; line-height: 1.8;">
Even one !important on a spacing property causes a failure.
</p>
Correct: Inline styles without !important
These examples pass because users can override the inline values with a custom stylesheet.
<!-- line-height without !important — overridable -->
<p style="line-height: 1.5;">
Users can adjust this line height with a custom stylesheet.
</p>
<!-- letter-spacing without !important — overridable -->
<p style="letter-spacing: 2px;">
Users can adjust this letter spacing.
</p>
<!-- word-spacing without !important — overridable -->
<p style="word-spacing: 4px;">
Users can adjust this word spacing.
</p>
<!-- Multiple spacing properties, all without !important -->
<p style="word-spacing: 4px; letter-spacing: 2px; line-height: 1.8;">
All three spacing properties can be overridden by the user.
</p>
<!-- font-size with !important is fine — not a text-spacing property -->
<p style="font-size: 200%;">
This does not trigger the rule.
</p>
Best practice: Use an external stylesheet instead
<!-- HTML -->
<p class="readable-text">
Styles are defined in the stylesheet, giving users full control.
</p>
/* CSS */
.readable-text {
line-height: 1.8;
letter-spacing: 0.05em;
word-spacing: 0.1em;
}
By keeping text-spacing styles in a stylesheet, you make it easy for users to apply their own overrides while maintaining your default design.
When a button lacks an accessible name, assistive technologies like screen readers can only announce it generically — for example, as “button” — with no indication of its purpose. This is a critical barrier for people who are blind or deafblind, as they rely entirely on programmatically determined names to understand and interact with interface controls. A sighted user might infer a button’s purpose from an icon or visual context, but without a text-based name, that information is completely lost to assistive technology users.
This rule maps to WCAG 2.0, 2.1, and 2.2 Success Criterion 4.1.2: Name, Role, Value (Level A), which requires that all user interface components have a name that can be programmatically determined. It is also covered by Section 508, EN 301 549 (9.4.1.2), and Trusted Tester guidelines, which require that the purpose of every link and button be determinable from its accessible name, description, or context.
How to fix it
Ensure every <button> element or element with role="button" has an accessible name through one of these methods:
- Visible text content inside the button element.
-
A non-empty
aria-labelattribute that describes the button’s purpose. -
An
aria-labelledbyattribute that references an element containing visible, non-empty text. -
A
titleattribute (use as a last resort, sincetitletooltips are inconsistently exposed across devices).
If a button is purely decorative and should be hidden from assistive technologies, you can assign role="presentation" or role="none" and remove it from the tab order with tabindex="-1". However, this is rare for interactive buttons.
Common mistakes to avoid
-
Leaving a
<button>element completely empty. -
Using only a
valueattribute on a<button>— unlike<input>elements, thevalueattribute on<button>does not provide an accessible name. -
Setting
aria-labelto an empty string (aria-label=""). -
Pointing
aria-labelledbyto an element that doesn’t exist or contains no text. - Using only an icon or image inside a button without providing alternative text.
Examples
Incorrect: empty button
<button id="search"></button>
A screen reader announces this as “button” with no indication of its purpose.
Incorrect: button with only a value attribute
<button id="submit" value="Submit"></button>
The value attribute does not set the accessible name for <button> elements.
Incorrect: empty aria-label
<button id="close" aria-label=""></button>
An empty aria-label results in no accessible name.
Incorrect: aria-labelledby pointing to a missing or empty element
<button id="save" aria-labelledby="save-label"></button>
<div id="save-label"></div>
The referenced element exists but contains no text, so the button has no accessible name.
Correct: button with visible text
<button>Submit order</button>
Correct: icon button with aria-label
<button aria-label="Close dialog">
<svg aria-hidden="true" focusable="false">
<use href="#icon-close"></use>
</svg>
</button>
The aria-label provides the accessible name, while aria-hidden="true" on the SVG prevents duplicate announcements.
Correct: button labeled by another element
<h2 id="section-title">Shopping cart</h2>
<button aria-labelledby="section-title">
<svg aria-hidden="true" focusable="false">
<use href="#icon-arrow"></use>
</svg>
</button>
The button’s accessible name is drawn from the referenced heading text.
Correct: button with aria-label and visible text
<button aria-label="Search products">Search</button>
When both aria-label and inner text are present, aria-label takes precedence as the accessible name. Use this when you need a more descriptive name than what the visible text alone conveys.
Correct: button with title (last resort)
<button title="Print this page">
<svg aria-hidden="true" focusable="false">
<use href="#icon-print"></use>
</svg>
</button>
The title attribute provides an accessible name, but visible text or aria-label are preferred because title tooltips may not be available to touch-screen or keyboard-only users.
Websites typically repeat navigation links, branding, and other interface elements across every page. While sighted mouse users can visually scan past these blocks and click wherever they want, keyboard-only users and screen reader users must navigate through every interactive element sequentially. Without a bypass mechanism, a keyboard user might need to press Tab dozens of times just to reach the primary content on each new page they visit. For users with severe motor impairments, this can take several minutes per page and cause fatigue or physical pain. Even users with less severe limitations will experience frustrating delays compared to mouse users, who can reach any link in a second or two.
Screen reader users also benefit significantly from bypass mechanisms. Landmarks like <main>, <nav>, and <header> allow screen readers to present a structural outline of the page, enabling users to jump directly to the section they need. Headings serve a similar purpose — screen reader users can navigate by heading level to quickly locate the main content area.
This rule maps to WCAG 2.4.1 Bypass Blocks (Level A), which requires that a mechanism is available to bypass blocks of content repeated on multiple pages. It is also required by Section 508 (specifically §1194.22(o)), the Trusted Tester guidelines, and EN 301 549. Because it is a Level A requirement, it represents the minimum baseline for accessibility compliance.
How the Rule Works
The axe bypass rule checks that a page includes at least one of the following:
-
A landmark region (such as
<main>,<nav>,<header>, or<footer>) -
A heading (an
<h1>through<h6>element) - An internal skip link (an anchor link that points to a location further down the page)
If none of these are present, the rule flags the page as a failure.
How to Fix It
The best approach is to use HTML5 landmark elements to structure your page. At a minimum, include a <main> element that wraps the primary content of the page. You should also use <header>, <nav>, and <footer> to identify other common sections. A page should have only one <main> landmark.
Additionally, consider adding a skip navigation link as the very first focusable element on the page. This provides an immediate shortcut for keyboard users who don’t use screen readers and may not be able to navigate by landmarks.
Prefer native HTML5 elements over their ARIA equivalents. For example, use <main> rather than <div role="main">. Native elements are better supported and require less code.
Examples
Incorrect: No Landmarks, Headings, or Skip Links
This page has no structural landmarks, no headings, and no skip link. Keyboard users must tab through every element to reach the content.
<div class="header">
<div class="logo">My Site</div>
<div class="nav">
<a href="/home">Home</a>
<a href="/about">About</a>
<a href="/contact">Contact</a>
</div>
</div>
<div class="content">
<p>This is the main content of the page.</p>
</div>
<div class="footer">
<p>Footer information</p>
</div>
Correct: Using HTML5 Landmark Elements
Replacing generic <div> wrappers with semantic HTML5 elements gives the page proper structure that assistive technologies can use for navigation.
<header>
<div class="logo">My Site</div>
<nav>
<a href="/home">Home</a>
<a href="/about">About</a>
<a href="/contact">Contact</a>
</nav>
</header>
<main>
<h1>Welcome</h1>
<p>This is the main content of the page.</p>
<section>
<h2>Latest News</h2>
<p>Section content here.</p>
</section>
</main>
<footer>
<p>Footer information</p>
</footer>
Correct: Adding a Skip Navigation Link
A skip link gives keyboard users an immediate way to bypass repeated content. It is typically visually hidden until it receives focus.
<body>
<a class="skip-link" href="#main-content">Skip to main content</a>
<header>
<nav>
<a href="/home">Home</a>
<a href="/about">About</a>
<a href="/contact">Contact</a>
</nav>
</header>
<main id="main-content">
<h1>Page Title</h1>
<p>This is the main content of the page.</p>
</main>
<footer>
<p>Footer information</p>
</footer>
</body>
.skip-link {
position: absolute;
left: -9999px;
top: auto;
width: 1px;
height: 1px;
overflow: hidden;
}
.skip-link:focus {
position: static;
width: auto;
height: auto;
overflow: visible;
}
When the skip link receives keyboard focus, it becomes visible, and pressing Enter moves focus directly to the <main> element. Combined with proper landmark elements, this gives all users fast, reliable access to the page’s primary content.