diff --git a/cid/cid.py b/cid/cid.py
index 07b1bed..6f8ad8b 100644
--- a/cid/cid.py
+++ b/cid/cid.py
@@ -255,12 +255,15 @@ def buffer(self) -> bytes:
         """
         return b"".join([bytes([self.version]), multicodec.add_prefix(self.codec, self.multihash)])
 
-    def encode(self, encoding: str | None = "base58btc") -> bytes:
+    def encode(self, encoding: str | None = "base32") -> bytes:
         """
-        Encoded version of the raw representation
+        Encoded version of the raw representation.
+
+        The default encoding is ``base32`` (CID specification default for CIDv1).
+        Callers that need the previous default can pass ``"base58btc"`` explicitly.
 
         :param str encoding: the encoding to use to encode the raw representation,
-            should be supported by ``py-multibase``
+            should be supported by ``py-multibase``. Defaults to ``base32``.
         :return: encoded raw representation with the given encoding
         :rtype: bytes
         """
diff --git a/newsfragments/66.breaking.rst b/newsfragments/66.breaking.rst
new file mode 100644
index 0000000..4390717
--- /dev/null
+++ b/newsfragments/66.breaking.rst
@@ -0,0 +1,3 @@
+Change CIDv1.encode() default encoding from base58btc to base32 to conform with the CID specification.
+
+This also changes ``str(cid)`` for CIDv1 objects (which uses ``encode()``), and therefore IPLD JSON serialization via ``to_json_dict()`` / ``CIDJSONEncoder``, which emit the string form. Callers that require base58btc must pass ``encode("base58btc")`` explicitly.
diff --git a/tests/test_cid.py b/tests/test_cid.py
index 9554eb0..da7b6f3 100644
--- a/tests/test_cid.py
+++ b/tests/test_cid.py
@@ -81,8 +81,8 @@ def test_buffer(self, cid):
         assert buffer[1:] == multicodec.add_prefix(self.TEST_CODEC, cid.multihash)
 
     def test_encode_default(self, cid):
-        """#encode defaults to base58btc encoding"""
-        assert cid.encode() == b"zdj7WhuEjrB52m1BisYCtmjH1hSKa7yZ3jEZ9JcXaFRD51wVz"
+        """#encode defaults to base32 encoding"""
+        assert cid.encode() == b"bafybeifzjut3te2nhyekklss27nh3k72ysco7y32koao5eei66wof36n5e"
 
     @pytest.mark.parametrize("codec", ENCODINGS)
     def test_encode_encoding(self, cid, codec):