diff --git a/ddtrace/contrib/kafka/patch.py b/ddtrace/contrib/kafka/patch.py
index 4218b11b4c2..0d9c6860031 100644
--- a/ddtrace/contrib/kafka/patch.py
+++ b/ddtrace/contrib/kafka/patch.py
@@ -244,7 +244,12 @@ def traced_poll(func, instance, args, kwargs):
                 ):
                     span.set_tag_str(kafkax.MESSAGE_KEY, message_key)
                 span.set_tag(kafkax.PARTITION, message.partition())
-                span.set_tag_str(kafkax.TOMBSTONE, str(len(message) == 0))
+                is_tombstone = False
+                try:
+                    is_tombstone = len(message) == 0
+                except TypeError:  # https://github.com/confluentinc/confluent-kafka-python/issues/1192
+                    pass
+                span.set_tag_str(kafkax.TOMBSTONE, str(is_tombstone))
                 span.set_tag(kafkax.MESSAGE_OFFSET, message_offset)
             span.set_tag(SPAN_MEASURED_KEY)
             rate = config.kafka.get_analytics_sample_rate()
diff --git a/releasenotes/notes/kafka-typeerror-b39a35c5338b05aa.yaml b/releasenotes/notes/kafka-typeerror-b39a35c5338b05aa.yaml
new file mode 100644
index 00000000000..e759637f2b2
--- /dev/null
+++ b/releasenotes/notes/kafka-typeerror-b39a35c5338b05aa.yaml
@@ -0,0 +1,5 @@
+---
+fixes:
+  - |
+    kafka: This fix resolves an issue where the use of a Kafka ``DeserializingConsumer`` could result in
+    a crash when the deserializer in use returns a type without a ``__len__`` attribute.
diff --git a/tests/contrib/kafka/test_kafka.py b/tests/contrib/kafka/test_kafka.py
index f5193764cdd..a3c86fa4a61 100644
--- a/tests/contrib/kafka/test_kafka.py
+++ b/tests/contrib/kafka/test_kafka.py
@@ -837,7 +837,7 @@ def json_deserializer(as_bytes, ctx):
         try:
             return json.loads(as_bytes)
         except json.decoder.JSONDecodeError:
-            return as_bytes
+            return  # return a type that has no __len__ because such types caused a crash at one point
 
     conf = {
         "bootstrap.servers": BOOTSTRAP_SERVERS,