These selectors enable the selection of an element based on the presence of an attribute alone (for example href ), or on various different matches against the value of the attribute.

Selector	Example	Description
[attr ]	a[title]	Matches elements with an attr attribute (whose name is the value in square brackets).
[attr =value ]	a[href="https://example.com"]	Matches elements with an attr attribute whose value is exactly value — the string inside the quotes.
[attr ~=value ]	p[class~="special"]	Matches elements with an attr attribute whose value is exactly value , or contains value in its (space separated) list of values.
[attr |=value ]	div[lang|="zh"]	Matches elements with an attr attribute whose value is exactly value or begins with value immediately followed by a hyphen.

In the example below you can see these selectors being used.

• By using li[class] we can match any list item with a class attribute. This matches all of the list items except the first one.
• li[class="a"] matches a selector with a class of a , but not a selector with a class of a with another space-separated class as part of the value. It selects the second list item.
• li[class~="a"] will match a class of a but also a value that contains the class of a as part of a whitespace-separated list. It selects the second and third list items.

These selectors allow for more advanced matching of substrings inside the value of your attribute. For example, if you had classes of box-warning and box-error and wanted to match everything that started with the string "box-", you could use [class^="box-"] to select them both (or [class|="box"] as described in section above).

Selector	Example	Description
[attr^=value]	li[class^="box-"]	Matches elements with an attr attribute, whose value begins with value .
[attr$=value]	li[class$="-box"]	Matches elements with an attr attribute whose value ends with value .
[attr*=value]	li[class*="box"]	Matches elements with an attr attribute whose value contains value anywhere within the string.

(Aside: It may help to note that ^ and $ have long been used as anchors in so-called regular expressions to mean begins with and ends with respectively.)

The next example shows usage of these selectors:

• li[class^="a"] matches any attribute value which starts with a , so matches the first two list items.
• li[class$="a"] matches any attribute value that ends with a , so matches the first and third list item.
• li[class*="a"] matches any attribute value where a appears anywhere in the string, so it matches all of our list items.

If you want to match attribute values case-insensitively you can use the value i before the closing bracket. This flag tells the browser to match ASCII characters case-insensitively. Without the flag the values will be matched according to the case-sensitivity of the document language — in HTML's case it will be case sensitive.

In the example below, the first selector will match a value that begins with a — it only matches the first list item because the other two list items start with an uppercase A. The second selector uses the case-insensitive flag and so matches all of the list items.

Now that we are done with attribute selectors, you can continue on to the next article and read about pseudo-class and pseudo-element selectors .