Throttle Function Calls in JavaScript

If you have written a search component that returns matching result as you type – you would realise that sending a request to the server on every key press is not required as that can result in burdening the server with too many requests.

A good practice is to throttle your requests to fire on a given amount time. Most users these days can type there name in 1/10 of a second, so 100-250ms is a good amount of delay before sending a request.


A JavaScript function written to throttle execution is called a debounce function. Or in another words, it limits the rate at which a function can fire.

Coding a debounce function

Let’s start by creating an index.html and adding an input field to it for search input.

<!DOCTYPE html>
<html lang="en">
  <meta charset="UTF-8">
  <title>Throttle function calls</title>
    <input type="text" id="search" placeholder="search">

Create a JavaScript file for our debounce function, let’s name it throttle.js.

function debounce(fn, delay) {
  var timer = null,
    delay = delay || 250;

	return function () {
		var args = arguments;


		timer = setTimeout(function() {
			fn.apply(this, args);
		}.bind(this), delay);

We can call this function inside an event handler.

var searchBox = document.getElementById('search');

searchBox.onkeyup = debounce(function(event) {
    console.log('searching .. ' + searchBox.value);

The returned function will be able to reference timer due to closure. Each call to the returned function will share this common timer.

var timer = null,

Calling debounce returns a new anonymous function.

return function () {

clearTimeout is the basic debounce behaviour where you can call this function several times, but it will only execute once before or after imposing a delay.
Each time the returned function is called, the timer starts over.


Now you can call the original function with apply. Apply lets you define the this object as well as the arguments (both captured before setTimeout).

The source code for this post can be found on Github.

Be first to comment

Leave a Reply