diff --git a/README.md b/README.md
index 13012c8..b024446 100644
--- a/README.md
+++ b/README.md
@@ -1,46 +1,28 @@
-Join all arguments together and normalize the resulting url.
+Join all arguments together and normalize the resulting URL.
 
 ## Install
 
-~~~
+```bash
 npm install url-join
-~~~
+```
+
+If you want to use it directly in a browser use a CDN like [Skypack](https://www.skypack.dev/view/url-join).
 
 ## Usage
 
-~~~javascript
-var urljoin = require('url-join');
+```javascript
+import urlJoin from 'url-join';
 
-var fullUrl = urljoin('http://www.google.com', 'a', '/b/cd', '?foo=123');
+const fullUrl = urlJoin('http://www.google.com', 'a', '/b/cd', '?foo=123');
 
 console.log(fullUrl);
-
-~~~
+```
 
 Prints:
 
-~~~
+```
 'http://www.google.com/a/b/cd?foo=123'
-~~~
-
-## Browser and AMD
-
-It also works in the browser, you can either include ```lib/url-join.js``` in your page:
-
-~~~html
-<script src="url-join.js"></script>
-<script type="text/javascript">
-	urljoin('http://blabla.com', 'foo?a=1')
-</script>
-~~~
-
-Or using an AMD module system like requirejs:
-
-~~~javascript
-define(['path/url-join.js'], function (urljoin) {
-  urljoin('http://blabla.com', 'foo?a=1');
-});
-~~~
+```
 
 ## License
 
diff --git a/bin/changelog b/bin/changelog.js
similarity index 96%
rename from bin/changelog
rename to bin/changelog.js
index 32bd741..73771a0 100755
--- a/bin/changelog
+++ b/bin/changelog.js
@@ -1,6 +1,6 @@
 #!/usr/bin/env node
+import changelog from 'conventional-changelog';
 
-var changelog = require('conventional-changelog');
 var semver_regex = /\bv?(?:0|[1-9][0-9]*)\.(?:0|[1-9][0-9]*)\.(?:0|[1-9][0-9]*)(?:-[\da-z\-]+(?:\.[\da-z\-]+)*)?(?:\+[\da-z\-]+(?:\.[\da-z\-]+)*)?\b/ig;
 
 const commitPartial = ` - {{header}}
diff --git a/lib/url-join.d.ts b/lib/url-join.d.ts
index 9fb12f2..317c2ba 100644
--- a/lib/url-join.d.ts
+++ b/lib/url-join.d.ts
@@ -6,4 +6,4 @@
 declare function urlJoin(...parts: string[]): string;
 declare function urlJoin(parts: string[]): string;
  
-export = urlJoin;
+export default urlJoin;
diff --git a/lib/url-join.js b/lib/url-join.js
index e23bb16..37f1303 100644
--- a/lib/url-join.js
+++ b/lib/url-join.js
@@ -1,78 +1,70 @@
-(function (name, context, definition) {
-  if (typeof module !== 'undefined' && module.exports) module.exports = definition();
-  else if (typeof define === 'function' && define.amd) define(definition);
-  else context[name] = definition();
-})('urljoin', this, function () {
-
-  function normalize (strArray) {
-    var resultArray = [];
-    if (strArray.length === 0) { return ''; }
-
-    if (typeof strArray[0] !== 'string') {
-      throw new TypeError('Url must be a string. Received ' + strArray[0]);
-    }
-
-    // If the first part is a plain protocol, we combine it with the next part.
-    if (strArray[0].match(/^[^/:]+:\/*$/) && strArray.length > 1) {
-      var first = strArray.shift();
-      strArray[0] = first + strArray[0];
-    }
+function normalize (strArray) {
+  var resultArray = [];
+  if (strArray.length === 0) { return ''; }
 
-    // There must be two or three slashes in the file protocol, two slashes in anything else.
-    if (strArray[0].match(/^file:\/\/\//)) {
-      strArray[0] = strArray[0].replace(/^([^/:]+):\/*/, '$1:///');
-    } else {
-      strArray[0] = strArray[0].replace(/^([^/:]+):\/*/, '$1://');
-    }
+  if (typeof strArray[0] !== 'string') {
+    throw new TypeError('Url must be a string. Received ' + strArray[0]);
+  }
 
-    for (var i = 0; i < strArray.length; i++) {
-      var component = strArray[i];
+  // If the first part is a plain protocol, we combine it with the next part.
+  if (strArray[0].match(/^[^/:]+:\/*$/) && strArray.length > 1) {
+    var first = strArray.shift();
+    strArray[0] = first + strArray[0];
+  }
 
-      if (typeof component !== 'string') {
-        throw new TypeError('Url must be a string. Received ' + component);
-      }
+  // There must be two or three slashes in the file protocol, two slashes in anything else.
+  if (strArray[0].match(/^file:\/\/\//)) {
+    strArray[0] = strArray[0].replace(/^([^/:]+):\/*/, '$1:///');
+  } else {
+    strArray[0] = strArray[0].replace(/^([^/:]+):\/*/, '$1://');
+  }
 
-      if (component === '') { continue; }
+  for (var i = 0; i < strArray.length; i++) {
+    var component = strArray[i];
 
-      if (i > 0) {
-        // Removing the starting slashes for each component but the first.
-        component = component.replace(/^[\/]+/, '');
-      }
-      if (i < strArray.length - 1) {
-        // Removing the ending slashes for each component but the last.
-        component = component.replace(/[\/]+$/, '');
-      } else {
-        // For the last component we will combine multiple slashes to a single one.
-        component = component.replace(/[\/]+$/, '/');
-      }
+    if (typeof component !== 'string') {
+      throw new TypeError('Url must be a string. Received ' + component);
+    }
 
-      resultArray.push(component);
+    if (component === '') { continue; }
 
+    if (i > 0) {
+      // Removing the starting slashes for each component but the first.
+      component = component.replace(/^[\/]+/, '');
+    }
+    if (i < strArray.length - 1) {
+      // Removing the ending slashes for each component but the last.
+      component = component.replace(/[\/]+$/, '');
+    } else {
+      // For the last component we will combine multiple slashes to a single one.
+      component = component.replace(/[\/]+$/, '/');
     }
 
-    var str = resultArray.join('/');
-    // Each input component is now separated by a single slash except the possible first plain protocol part.
+    resultArray.push(component);
 
-    // remove trailing slash before parameters or hash
-    str = str.replace(/\/(\?|&|#[^!])/g, '$1');
+  }
 
-    // replace ? in parameters with &
-    var parts = str.split('?');
-    str = parts.shift() + (parts.length > 0 ? '?': '') + parts.join('&');
+  var str = resultArray.join('/');
+  // Each input component is now separated by a single slash except the possible first plain protocol part.
 
-    return str;
-  }
+  // remove trailing slash before parameters or hash
+  str = str.replace(/\/(\?|&|#[^!])/g, '$1');
 
-  return function () {
-    var input;
+  // replace ? in parameters with &
+  var parts = str.split('?');
+  str = parts.shift() + (parts.length > 0 ? '?': '') + parts.join('&');
 
-    if (typeof arguments[0] === 'object') {
-      input = arguments[0];
-    } else {
-      input = [].slice.call(arguments);
-    }
+  return str;
+}
 
-    return normalize(input);
-  };
+export default function urlJoin() {
+  var input;
+
+  if (typeof arguments[0] === 'object') {
+    input = arguments[0];
+  } else {
+    input = [].slice.call(arguments);
+  }
 
-});
+  return normalize(input);
+}
diff --git a/package.json b/package.json
index 34ca8fd..135d38a 100644
--- a/package.json
+++ b/package.json
@@ -2,9 +2,14 @@
   "name": "url-join",
   "version": "4.0.1",
   "description": "Join urls and normalize as in path.join.",
-  "main": "lib/url-join.js",
-  "types": "lib/url-join.d.ts",
+  "type": "module",
+  "main": "./lib/url-join.js",
+  "exports": "./lib/url-join.js",
+  "types": "./lib/url-join.d.ts",
   "sideEffects": false,
+  "engines": {
+    "node": "^12.20.0 || ^14.13.1 || >=16.0.0"
+  },
   "scripts": {
     "test": "mocha --require should"
   },
diff --git a/test/tests.js b/test/tests.js
index 6876276..7e3dc48 100644
--- a/test/tests.js
+++ b/test/tests.js
@@ -1,151 +1,151 @@
-var urljoin = require('../lib/url-join');
-var assert = require('assert');
+import assert from 'assert';
+import urlJoin from '../lib/url-join.js';
 
 describe('url join', function () {
   it('should work for simple case', function () {
-    urljoin('http://www.google.com/', 'foo/bar', '?test=123')
+    urlJoin('http://www.google.com/', 'foo/bar', '?test=123')
       .should.eql('http://www.google.com/foo/bar?test=123');
   });
 
   it('should work for simple case with new syntax', function () {
-    urljoin(['http://www.google.com/', 'foo/bar', '?test=123'])
+    urlJoin(['http://www.google.com/', 'foo/bar', '?test=123'])
       .should.eql('http://www.google.com/foo/bar?test=123');
   });
 
   it('should work for hashbang urls', function () {
-    urljoin(['http://www.google.com', '#!', 'foo/bar', '?test=123'])
+    urlJoin(['http://www.google.com', '#!', 'foo/bar', '?test=123'])
       .should.eql('http://www.google.com/#!/foo/bar?test=123');
   });
 
   it('should be able to join protocol', function () {
-    urljoin('http:', 'www.google.com/', 'foo/bar', '?test=123')
+    urlJoin('http:', 'www.google.com/', 'foo/bar', '?test=123')
       .should.eql('http://www.google.com/foo/bar?test=123');
   });
 
   it('should be able to join protocol with slashes', function () {
-    urljoin('http://', 'www.google.com/', 'foo/bar', '?test=123')
+    urlJoin('http://', 'www.google.com/', 'foo/bar', '?test=123')
       .should.eql('http://www.google.com/foo/bar?test=123');
   });
 
   it('should remove extra slashes', function () {
-    urljoin('http:', 'www.google.com///', 'foo/bar', '?test=123')
+    urlJoin('http:', 'www.google.com///', 'foo/bar', '?test=123')
       .should.eql('http://www.google.com/foo/bar?test=123');
   });
 
   it('should not remove extra slashes in an encoded URL', function () {
-    urljoin('http:', 'www.google.com///', 'foo/bar', '?url=http%3A//Ftest.com')
+    urlJoin('http:', 'www.google.com///', 'foo/bar', '?url=http%3A//Ftest.com')
       .should.eql('http://www.google.com/foo/bar?url=http%3A//Ftest.com');
 
-    urljoin('http://a.com/23d04b3/', '/b/c.html')
+    urlJoin('http://a.com/23d04b3/', '/b/c.html')
       .should.eql('http://a.com/23d04b3/b/c.html')
       .should.not.eql('http://a.com/23d04b3//b/c.html');
   });
 
   it('should support anchors in urls', function () {
-    urljoin('http:', 'www.google.com///', 'foo/bar', '?test=123', '#faaaaa')
+    urlJoin('http:', 'www.google.com///', 'foo/bar', '?test=123', '#faaaaa')
       .should.eql('http://www.google.com/foo/bar?test=123#faaaaa');
   });
 
   it('should support protocol-relative urls', function () {
-    urljoin('//www.google.com', 'foo/bar', '?test=123')
+    urlJoin('//www.google.com', 'foo/bar', '?test=123')
       .should.eql('//www.google.com/foo/bar?test=123')
   });
 
   it('should support file protocol urls', function () {
-    urljoin('file:/', 'android_asset', 'foo/bar')
+    urlJoin('file:/', 'android_asset', 'foo/bar')
       .should.eql('file://android_asset/foo/bar')
 
-    urljoin('file:', '/android_asset', 'foo/bar')
+    urlJoin('file:', '/android_asset', 'foo/bar')
       .should.eql('file://android_asset/foo/bar')
   });
 
   it('should support absolute file protocol urls', function () {
-    urljoin('file:', '///android_asset', 'foo/bar')
+    urlJoin('file:', '///android_asset', 'foo/bar')
       .should.eql('file:///android_asset/foo/bar')
 
-    urljoin('file:///', 'android_asset', 'foo/bar')
+    urlJoin('file:///', 'android_asset', 'foo/bar')
       .should.eql('file:///android_asset/foo/bar')
 
-    urljoin('file:///', '//android_asset', 'foo/bar')
+    urlJoin('file:///', '//android_asset', 'foo/bar')
       .should.eql('file:///android_asset/foo/bar')
 
-    urljoin('file:///android_asset', 'foo/bar')
+    urlJoin('file:///android_asset', 'foo/bar')
       .should.eql('file:///android_asset/foo/bar')
   });
 
   it('should merge multiple query params properly', function () {
-    urljoin('http:', 'www.google.com///', 'foo/bar', '?test=123', '?key=456')
+    urlJoin('http:', 'www.google.com///', 'foo/bar', '?test=123', '?key=456')
       .should.eql('http://www.google.com/foo/bar?test=123&key=456');
 
-    urljoin('http:', 'www.google.com///', 'foo/bar', '?test=123', '?boom=value', '&key=456')
+    urlJoin('http:', 'www.google.com///', 'foo/bar', '?test=123', '?boom=value', '&key=456')
       .should.eql('http://www.google.com/foo/bar?test=123&boom=value&key=456');
 
-    urljoin('http://example.org/x', '?a=1', '?b=2', '?c=3', '?d=4')
+    urlJoin('http://example.org/x', '?a=1', '?b=2', '?c=3', '?d=4')
       .should.eql('http://example.org/x?a=1&b=2&c=3&d=4');
   });
 
   it('should merge slashes in paths correctly', function () {
-    urljoin('http://example.org', 'a//', 'b//', 'A//', 'B//')
+    urlJoin('http://example.org', 'a//', 'b//', 'A//', 'B//')
       .should.eql('http://example.org/a/b/A/B/');
   });
 
   it('should merge colons in paths correctly', function () {
-    urljoin('http://example.org/', ':foo:', 'bar')
+    urlJoin('http://example.org/', ':foo:', 'bar')
       .should.eql('http://example.org/:foo:/bar');
   });
 
   it('should merge just a simple path without URL correctly', function() {
-    urljoin('/', 'test')
+    urlJoin('/', 'test')
       .should.eql('/test');
   });
 
   it('should fail with segments that are not string', function() {
-    assert.throws(() => urljoin(true),
+    assert.throws(() => urlJoin(true),
                   /Url must be a string. Received true/);
-    assert.throws(() => urljoin('http://blabla.com/', 1),
+    assert.throws(() => urlJoin('http://blabla.com/', 1),
                   /Url must be a string. Received 1/);
-    assert.throws(() => urljoin('http://blabla.com/', undefined, 'test'),
+    assert.throws(() => urlJoin('http://blabla.com/', undefined, 'test'),
                   /Url must be a string. Received undefined/);
-    assert.throws(() => urljoin('http://blabla.com/', null, 'test'),
+    assert.throws(() => urlJoin('http://blabla.com/', null, 'test'),
                   /Url must be a string. Received null/);
-    assert.throws(() => urljoin('http://blabla.com/', { foo: 123 }, 'test'),
+    assert.throws(() => urlJoin('http://blabla.com/', { foo: 123 }, 'test'),
                   /Url must be a string. Received \[object Object\]/);
   });
 
   it('should merge a path with colon properly', function(){
-    urljoin('/users/:userId', '/cars/:carId')
+    urlJoin('/users/:userId', '/cars/:carId')
       .should.eql('/users/:userId/cars/:carId');
   });
 
   it('should merge slashes in protocol correctly', function () {
-    urljoin('http://example.org', 'a')
+    urlJoin('http://example.org', 'a')
       .should.eql('http://example.org/a');
-    urljoin('http:', '//example.org', 'a')
+    urlJoin('http:', '//example.org', 'a')
       .should.eql('http://example.org/a');
-    urljoin('http:///example.org', 'a')
+    urlJoin('http:///example.org', 'a')
       .should.eql('http://example.org/a');
-    urljoin('file:///example.org', 'a')
+    urlJoin('file:///example.org', 'a')
       .should.eql('file:///example.org/a');
 
-    urljoin('file:example.org', 'a')
+    urlJoin('file:example.org', 'a')
       .should.eql('file://example.org/a');
 
-    urljoin('file:/', 'example.org', 'a')
+    urlJoin('file:/', 'example.org', 'a')
       .should.eql('file://example.org/a');
-    urljoin('file:', '/example.org', 'a')
+    urlJoin('file:', '/example.org', 'a')
       .should.eql('file://example.org/a');
-    urljoin('file:', '//example.org', 'a')
+    urlJoin('file:', '//example.org', 'a')
       .should.eql('file://example.org/a');
   });
 
   it('should skip empty strings', function() {
-    urljoin('http://foobar.com', '', 'test')
+    urlJoin('http://foobar.com', '', 'test')
       .should.eql('http://foobar.com/test');
-    urljoin('', 'http://foobar.com', '', 'test')
+    urlJoin('', 'http://foobar.com', '', 'test')
       .should.eql('http://foobar.com/test');
   });
 
   it('should return an empty string if no arguments are supplied', function() {
-    urljoin().should.eql('');
+    urlJoin().should.eql('');
   });
 });