diff --git a/src/static/js/pad_utils.js b/src/static/js/pad_utils.js
index 48f0624e9..fbb9f1637 100644
--- a/src/static/js/pad_utils.js
+++ b/src/static/js/pad_utils.js
@@ -60,10 +60,10 @@ const wordCharRegex = new RegExp(`[${[
 const urlRegex = (() => {
   // TODO: wordCharRegex matches many characters that are not permitted in URIs. Are they included
   // here as an attempt to support IRIs? (See https://tools.ietf.org/html/rfc3987.)
-  const urlChar = `[-:@_.,~%+/?=&#!;()$'*${wordCharRegex.source.slice(1, -1)}]`;
+  const urlChar = `[-:@_.,~%+/?=&#!;()\\[\\]$'*${wordCharRegex.source.slice(1, -1)}]`;
   // Matches a single character that should not be considered part of the URL if it is the last
   // character that matches urlChar.
-  const postUrlPunct = '[:.,;?!)\'*]';
+  const postUrlPunct = '[:.,;?!)\\]\'*]';
   // Schemes that must be followed by ://
   const withAuth = `(?:${[
     '(?:x-)?man',
diff --git a/src/tests/frontend/specs/urls_become_clickable.js b/src/tests/frontend/specs/urls_become_clickable.js
index f170a2a41..bb3f00b34 100644
--- a/src/tests/frontend/specs/urls_become_clickable.js
+++ b/src/tests/frontend/specs/urls_become_clickable.js
@@ -28,7 +28,7 @@ describe('urls', function () {
   });
 
   describe('special characters inside URL', function () {
-    for (const char of '-:@_.,~%+/?=&#!;()$\'*') {
+    for (const char of '-:@_.,~%+/?=&#!;()[]$\'*') {
       const url = `https://etherpad.org/${char}foo`;
       it(url, async function () {
         await helper.edit(url);
@@ -41,7 +41,7 @@ describe('urls', function () {
   });
 
   describe('punctuation after URL is ignored', function () {
-    for (const char of ':.,;?!)\'*]') {
+    for (const char of ':.,;?!)]\'*') {
       const want = 'https://etherpad.org';
       const input = want + char;
       it(input, async function () {
@@ -53,21 +53,4 @@ describe('urls', function () {
       });
     }
   });
-
-  // Square brackets are in the RFC3986 reserved set so they can legally appear in URIs, but they
-  // are explicitly excluded from linkification because including them is usually not desired (e.g.,
-  // it can interfere with wiki/markdown link syntax).
-  describe('square brackets are excluded from linkified URLs', function () {
-    for (const char of '[]') {
-      const want = 'https://etherpad.org/';
-      const input = `${want}${char}foo`;
-      it(input, async function () {
-        await helper.edit(input);
-        await helper.waitForPromise(() => txt().find('a').length === 1);
-        const link = txt().find('a');
-        expect(link.attr('href')).to.be(want);
-        expect(link.text()).to.be(want);
-      });
-    }
-  });
 });