diff --git a/Lib/encodings/idna.py b/Lib/encodings/idna.py
index d31ee07ab45b76..8e245342f23771 100644
--- a/Lib/encodings/idna.py
+++ b/Lib/encodings/idna.py
@@ -77,7 +77,8 @@ def ToASCII(label):  # type: (str) -> bytes
         if len(label) == 0:
             raise UnicodeEncodeError("idna", label, 0, 1, "label empty")
         else:
-            raise UnicodeEncodeError("idna", label, 0, len(label), "label too long")
+            raise UnicodeEncodeError("idna", label, 0, len(label),
+                                     f"label too long: {label!r}")
 
     # Step 2: nameprep
     label = nameprep(label)
@@ -95,7 +96,8 @@ def ToASCII(label):  # type: (str) -> bytes
         if len(label) == 0:
             raise UnicodeEncodeError("idna", label, 0, 1, "label empty")
         else:
-            raise UnicodeEncodeError("idna", label, 0, len(label), "label too long")
+            raise UnicodeEncodeError("idna", label, 0, len(label),
+                                     f"label too long: {label!r}")
 
     # Step 5: Check ACE prefix
     if label.lower().startswith(sace_prefix):
@@ -112,7 +114,8 @@ def ToASCII(label):  # type: (str) -> bytes
     # do not check for empty as we prepend ace_prefix.
     if len(label_ascii) < 64:
         return label_ascii
-    raise UnicodeEncodeError("idna", label, 0, len(label), "label too long")
+    raise UnicodeEncodeError("idna", label, 0, len(label),
+                             f"label too long: {label!r}")
 
 def ToUnicode(label):
     if len(label) > 1024:
@@ -201,7 +204,7 @@ def encode(self, input, errors='strict'):
                 if len(label) >= 64:
                     offset = sum(len(l) for l in labels[:i]) + i
                     raise UnicodeEncodeError("idna", input, offset, offset+len(label),
-                                             "label too long")
+                                             f"label too long: {label.decode('ascii')!r}")
             return result, len(input)
 
         result = bytearray()
diff --git a/Lib/test/test_codecs.py b/Lib/test/test_codecs.py
index c31faec9ee5214..cbfc26675308e5 100644
--- a/Lib/test/test_codecs.py
+++ b/Lib/test/test_codecs.py
@@ -1640,6 +1640,29 @@ def test_builtin_decode_length_limit(self):
         with self.assertRaisesRegex(UnicodeDecodeError, "too long"):
             (b"xn--016c"+b"a"*70).decode("idna")
 
+    def test_error_message_label_too_long(self):
+        # gh-53891: "label too long" errors should include the offending label
+        long_label = "a" * 70
+        domain = f"{long_label}.com"
+        with self.assertRaises(UnicodeEncodeError) as cm:
+            domain.encode("idna")
+        self.assertIn(long_label, cm.exception.reason)
+        self.assertIn("label too long", cm.exception.reason)
+
+    def test_error_message_label_too_long_non_ascii(self):
+        # gh-53891: non-ASCII labels should also be included in the message
+        long_label = "\xe4" * 70
+        domain = f"{long_label}.com"
+        with self.assertRaises(UnicodeEncodeError) as cm:
+            domain.encode("idna")
+        self.assertIn("label too long", cm.exception.reason)
+
+    def test_error_message_label_empty(self):
+        # gh-53891: empty label errors should have a clear reason
+        with self.assertRaises(UnicodeEncodeError) as cm:
+            "a..com".encode("idna")
+        self.assertEqual(cm.exception.reason, "label empty")
+
     def test_stream(self):
         r = codecs.getreader("idna")(io.BytesIO(b"abc"))
         r.read(3)
diff --git a/Misc/NEWS.d/next/Library/2026-02-15-23-25-41.gh-issue-53891.JlZccA.rst b/Misc/NEWS.d/next/Library/2026-02-15-23-25-41.gh-issue-53891.JlZccA.rst
new file mode 100644
index 00000000000000..8bd9a3fdfc0cbc
--- /dev/null
+++ b/Misc/NEWS.d/next/Library/2026-02-15-23-25-41.gh-issue-53891.JlZccA.rst
@@ -0,0 +1,2 @@
+Include the offending label value in IDNA codec "label too long" error
+messages to make them more informative and easier to debug.