# Potentially bad value X for attribute “rel” on element “a” Typo for Y?

> Canonical HTML version: https://rocketvalidator.com/html-validation/potentially-bad-value-x-for-attribute-rel-on-element-a-typo-for-y
> Attribution: Rocket Validator (https://rocketvalidator.com)
> License: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/)

The `rel` attribute on an `<a>` element contains a token that looks like a misspelled link type, such as `_nofollow` where `nofollow` was meant.

`rel` holds a space-separated list of keywords that describe how the current page relates to the linked resource. Browsers and search engines only act on the keywords they recognize, and they compare each token as written, ignoring case. A near miss such as `_nofollow`, `nofolow` or `no-follow` is not read as `nofollow`, so search engines see an ordinary followed link. A misspelled `noreferer` has a similar effect: the browser still sends the `Referer` header to the destination. The HTTP header is spelled with one "r", but the keyword is `noreferrer`.

The validator does not report every unknown token, because `rel` also accepts extension values defined outside the HTML standard. It warns only when a token is close to a known keyword, and the message ends with the keyword it suspects you meant. When the attribute holds several tokens, the message quotes the whole value, so check each token to find the misspelled one.

A leading underscore is a frequent cause. Keywords for the `target` attribute start with one (`_blank`, `_self`), and the habit carries over to `rel` on the same link. No `rel` keyword starts with an underscore.

## Invalid example

```html
<a href="https://example.com/" target="_blank" rel="_nofollow">Partner site</a>
<a href="https://example.org/" rel="noreferer">Another site</a>
```

## Valid example

```html
<a href="https://example.com/" target="_blank" rel="nofollow">Partner site</a>
<a href="https://example.org/" rel="noreferrer">Another site</a>
```