diff --git a/dataclasses_avroschema/model_generator/lang/python/base.py b/dataclasses_avroschema/model_generator/lang/python/base.py
index 48903801..2c59d124 100644
--- a/dataclasses_avroschema/model_generator/lang/python/base.py
+++ b/dataclasses_avroschema/model_generator/lang/python/base.py
@@ -327,6 +327,9 @@ def parse_logical_type(self, *, field: JsonDict) -> str:
         elif logical_type == field_utils.DECIMAL:
             # this is a special case for logical types
             type = self.parse_decimal(field=field, default=default)
+        elif logical_type not in self.logical_types_imports:
+            # Then it is a custom logicalType, so we default to the native type
+            type = self.get_language_type(type=field["type"])
         else:
             # add the logical type import
             self.imports.add(self.logical_types_imports[logical_type])
diff --git a/docs/logical_types.md b/docs/logical_types.md
index 72709361..c3d26ab3 100644
--- a/docs/logical_types.md
+++ b/docs/logical_types.md
@@ -13,6 +13,9 @@ The following list represent the avro logical types mapped to python types:
 | string    |  uuid        | uuid.UUID |
 | bytes     | decimal      | types.condecimal |
 
+!!! note
+    Custom `logicalTypes` are not supported
+
 ## Date
 
 ```python title="Date example"
diff --git a/docs/model_generator.md b/docs/model_generator.md
index 58fc72a2..8e80d913 100644
--- a/docs/model_generator.md
+++ b/docs/model_generator.md
@@ -123,6 +123,46 @@ class Address(AvroModel):
 
 Generating a single module from multiple schemas is useful for example to group schemas that belong to the same namespace.
 
+## LogicalTypes
+
+Native `logicalTypes` are supported by `dataclasses-avroschema` but custom ones are not. If you defined a custom `logicalType` then
+the fallback is used when generating the field. In the next example we have a `logicalType` defined as `url`, which is not a native one,
+then the model generated will use `string`
+
+```python
+from dataclasses_avroschema import ModelGenerator, ModelType
+
+model_generator = ModelGenerator()
+
+
+schema = {
+    "type": "record",
+    "name": "TestEvent",
+    "fields": [
+        {
+            "name": "regular",
+            "type": {
+                "type": "string",
+                "logicalType": "url"
+            },
+            "doc": "Urls"
+        }
+    ],
+}
+
+print(model_generator.render(schema=schema))
+
+"""
+from dataclasses_avroschema import AvroModel
+import dataclasses
+
+
+@dataclasses.dataclass
+class TestEvent(AvroModel):
+    regular: str = dataclasses.field(metadata={'doc': 'Urls'})
+"""
+```
+
 ## Render Pydantic models
 
 It is also possible to render `BaseModel` (pydantic) and `AvroBaseModel` (avro + pydantic) models as well.
diff --git a/tests/model_generator/conftest.py b/tests/model_generator/conftest.py
index ff3800db..f494f192 100644
--- a/tests/model_generator/conftest.py
+++ b/tests/model_generator/conftest.py
@@ -759,6 +759,26 @@ def schema_with_logical_types_field_order() -> JsonDict:
     }
 
 
+@pytest.fixture
+def schema_with_unknown_logical_types() -> JsonDict:
+    return {
+        "type": "record",
+        "name": "TestEvent",
+        "namespace": "com.example",
+        "fields": [
+            {"name": "occurredAt", "type": {"type": "long", "logicalType": "timestamp-millis"}, "doc": "Event time"},
+            {
+                "name": "previous",
+                "type": {
+                    "type": "record",
+                    "name": "urls",
+                    "fields": [{"name": "regular", "type": {"type": "string", "logicalType": "url"}, "doc": "Urls"}],
+                },
+            },
+        ],
+    }
+
+
 @pytest.fixture
 def schema_with_pydantic_fields() -> JsonDict:
     return {
diff --git a/tests/model_generator/test_model_generator.py b/tests/model_generator/test_model_generator.py
index c59fdea7..3c3d29bd 100644
--- a/tests/model_generator/test_model_generator.py
+++ b/tests/model_generator/test_model_generator.py
@@ -594,6 +594,36 @@ class LogicalTypes(AvroModel):
     assert result.strip() == expected_result.strip()
 
 
+def test_schema_with_unknown_logical_types(schema_with_unknown_logical_types: types.JsonDict) -> None:
+    expected_result = """
+from dataclasses_avroschema import AvroModel
+import dataclasses
+import datetime
+
+
+@dataclasses.dataclass
+class Urls(AvroModel):
+    regular: str = dataclasses.field(metadata={'doc': 'Urls'})
+
+    
+    class Meta:
+        schema_name = "urls"
+
+@dataclasses.dataclass
+class TestEvent(AvroModel):
+    occurredAt: datetime.datetime = dataclasses.field(metadata={'doc': 'Event time'})
+    previous: Urls
+
+    
+    class Meta:
+        namespace = "com.example"
+
+"""
+    model_generator = ModelGenerator()
+    result = model_generator.render(schema=schema_with_unknown_logical_types)
+    assert result.strip() == expected_result.strip()
+
+
 def test_field_order(schema_with_logical_types_field_order: types.JsonDict) -> None:
     release_datetime = render_datetime(value=1570903062000, format=field_utils.TIMESTAMP_MILLIS)
     release_datetime_micro = render_datetime(value=1570903062000000, format=field_utils.TIMESTAMP_MICROS)