Returns a static (not live) NodeList of all elements descended from the element on which it is invoked that matches the specified group of CSS selectors. (The base element itself is not included, even if it matches.)

Note: The definition of this API was moved to the ParentNode interface.


elementList = baseElement.querySelectorAll(selectors);


is a non-live node list [ NodeList[elements] of element objects.
is an element object.
is a group of selectors to match on or elements of the DOM. 


This example returns a list of all the p elements in the HTML document body:

let matches = document.body.querySelectorAll('p'); 

This example returns a list of p children elements under a container, whose parent is a div that has the class 'highlighted':

let el = document.querySelector('#test');    //return an element with id='test'
let matches = el.querySelectorAll('div.highlighted > p'); // return a NodeList of p wrapped in a div with attribute class "highlighted"

This example returns a list of iframe elements that contain a attribute 'data-src':

let matches = el.querySelectorAll('iframe[data-src]'); 


If the specified "selectors" are not found inside the DOM of the page, the method queryselectorAll returns an empty NodeList as specified below:

> let x = document.body.querySelectorAll('.highlighted'); //case: if the class highlighted doesn't exist in any attribute "class" of the DOM the result is
> [] //empty NodeList

querySelectorAll() was introduced in the WebApps API.

The string argument pass to querySelectorAll must follow the CSS syntax. See document.querySelector for a concrete example.

We could access a single item inside the NodeList in the following way:

let x = document.body.querySelectorAll('.highlighted'); 
x.length; //return the size of x
x[i_item]; //where i_item has a value between 0 and x.length-1. The operator "[]" return as in an array the element at index "i_item"

We could iterate inside a NodeList with the construct for(....) {...} as in the following code:

 let x = document.body.querySelectorAll('.highlighted');
 let index = 0;
 for( index=0; index < x.length; index++ ) {

So in the above way, it is possible to manage and modify the behaviour of the page.


querySelectorAll() behaves differently than most common JavaScript DOM libraries, which might lead to unexpected results:

<div class="outer">
  <div class="select">
    <div class="inner">
let select = document.querySelector('.select');
let inner = select.querySelectorAll('.outer .inner');
inner.length; // 1, not 0!

In this example, when selecting .outer .inner in the context of .select, .inner is still found, even though .outer is not a descendant of the baseElement (.select).
querySelectorAll() only verifies that the last element in the selector is within the baseElement.

The :scope pseudo-class restores the expected behavior, only matching selectors on descendants of the baseElement:

let select = document.querySelector('.select');
let inner = select.querySelectorAll(':scope .outer .inner');
inner.length; // 0


Specification Status Comment
Selectors API Level 1
The definition of 'querySelectorAll' in that specification.
Recommendation Initial definition

Browser compatibility

Feature Chrome Firefox (Gecko) Internet Explorer Opera Safari (WebKit)
Basic support 1 3.5 (1.9.1) 8 10 3.2 (525.3)
:scope pseudo-class (Yes) 32 No support 15[1] 7.0
Feature Android Firefox Mobile (Gecko) IE Phone Opera Mobile Safari Mobile
Basic support (Yes) 1.0 (1.9.1) ? ? (Yes)
:scope pseudo-class ? 32 No support No support 7.0

[1] Supported in Opera 15+ by enabling the "Enable <style scoped>" or "Enable experimental Web Platform features" flag in chrome://flags.

See also