diff --git a/jsf/parser.py b/jsf/parser.py
index 3174429..d8a6765 100644
--- a/jsf/parser.py
+++ b/jsf/parser.py
@@ -128,6 +128,11 @@ def __parse_object(
         self, name: str, path: str, schema: Dict[str, Any], root: Optional[AllTypes] = None
     ) -> Object:
         _, is_nullable = self.__is_field_nullable(schema)
+        schema_without_props = {
+            k: v
+            for k, v in schema.items()
+            if k not in ("properties", "patternProperties", "dependencies")
+        }
         model = Object.from_dict(
             {
                 "name": name,
@@ -135,7 +140,7 @@ def __parse_object(
                 "is_nullable": is_nullable,
                 "allow_none_optionals": self.allow_none_optionals,
                 "max_recursive_depth": self.max_recursive_depth,
-                **schema,
+                **schema_without_props,
             }
         )
         root = model if root is None else root
@@ -292,6 +297,18 @@ def __parse_definition(
                     isinstance(x, dict) for x in schema.get("items", [])
                 ):
                     return self.__parse_tuple(name, path, schema, root)
+                # arrays without an "items" definition should still be valid and
+                # simply produce empty lists
+                return Array.from_dict(
+                    {
+                        "name": name,
+                        "path": path,
+                        "is_nullable": is_nullable,
+                        "allow_none_optionals": self.allow_none_optionals,
+                        "max_recursive_depth": self.max_recursive_depth,
+                        **schema,
+                    }
+                )
             else:
                 return self.__parse_primitive(name, path, schema)
         elif "$ref" in schema:
diff --git a/jsf/schema_types/array.py b/jsf/schema_types/array.py
index 8686c52..ba1e20a 100644
--- a/jsf/schema_types/array.py
+++ b/jsf/schema_types/array.py
@@ -9,7 +9,12 @@
 class Array(BaseSchema):
     items: Optional[BaseSchema] = None
     contains: Optional[BaseSchema] = None  # NOTE: Validation only
-    minItems: Optional[int] = 0
+    # If `items` is provided in the schema, JSON Schema treats the array as
+    # having an item type.  In that case JSF should emit at least one element
+    # by default.  Using ``None`` here allows us to distinguish between the
+    # schema omitting ``minItems`` and explicitly setting ``minItems`` to ``0``
+    # which callers may rely on.
+    minItems: Optional[int] = None
     maxItems: Optional[int] = 5
     uniqueItems: Optional[bool] = False
     fixed: Optional[Union[int, str]] = Field(None, alias="$fixed")
@@ -22,33 +27,53 @@ def generate(self, context: Dict[str, Any]) -> Optional[List[Any]]:
         try:
             return super().generate(context)
         except ProviderNotSetException:
+            if self.items is None:
+                # No item schema means we cannot infer what the array should
+                # contain, therefore return an empty list.
+                return []
+
             if isinstance(self.fixed, str):
                 self.minItems = self.maxItems = eval(self.fixed, context)()
             elif isinstance(self.fixed, int):
                 self.minItems = self.maxItems = self.fixed
 
             depth = context["state"]["__depth__"]
+
+            # ``minItems`` may be ``None`` when it wasn't provided in the
+            # schema.  In that scenario we want non-empty arrays if an item
+            # schema exists.  When the user explicitly sets ``minItems`` to 0
+            # we honour that and allow empty arrays.
+            min_items = (
+                int(self.minItems)
+                if self.minItems is not None
+                else (0 if self.items is None else 1)
+            )
+            max_items = int(self.maxItems) if self.maxItems is not None else 5
+
             output = []
-            for _ in range(random.randint(int(self.minItems), int(self.maxItems))):
+            for _ in range(random.randint(min_items, max_items)):
                 output.append(self.items.generate(context))
                 context["state"]["__depth__"] = depth
             if self.uniqueItems and self.items.type == "object":
                 output = [dict(s) for s in {frozenset(d.items()) for d in output}]
-                while len(output) < self.minItems:
+                while len(output) < min_items:
                     output.append(self.items.generate(context))
                     output = [dict(s) for s in {frozenset(d.items()) for d in output}]
                     context["state"]["__depth__"] = depth
             elif self.uniqueItems:
                 output = set(output)
-                while len(output) < self.minItems:
+                while len(output) < min_items:
                     output.add(self.items.generate(context))
                     context["state"]["__depth__"] = depth
                 output = list(output)
             return output
 
     def model(self, context: Dict[str, Any]) -> Tuple[Type, Any]:
-        _type = eval(
-            f"List[Union[{','.join([self.items.model(context)[0].__name__])}]]",
-            context["__internal__"],
-        )
+        if self.items is None:
+            _type = List[Any]
+        else:
+            _type = eval(
+                f"List[Union[{','.join([self.items.model(context)[0].__name__])}]]",
+                context["__internal__"],
+            )
         return self.to_pydantic(context, _type)
diff --git a/jsf/schema_types/object.py b/jsf/schema_types/object.py
index 7e3fcf6..d8e19a3 100644
--- a/jsf/schema_types/object.py
+++ b/jsf/schema_types/object.py
@@ -19,14 +19,14 @@ class PropertyNames(BaseModel):
 
 
 class Object(BaseSchema):
-    properties: Dict[str, BaseSchema] = {}
+    properties: List[BaseSchema] = []
     additionalProperties: Optional[Union[bool, BaseSchema]] = None
     required: Optional[List[str]] = None
     propertyNames: Optional[PropertyNames] = None
     minProperties: Optional[int] = None
     maxProperties: Optional[int] = None
     dependencies: Optional[Union[PropertyDependency, SchemaDependency]] = None
-    patternProperties: Optional[Dict[str, BaseSchema]] = None
+    patternProperties: Optional[List[BaseSchema]] = None
 
     @classmethod
     def from_dict(cls, d: Dict[str, Any]) -> "Object":
diff --git a/jsf/tests/data/empty-list-pro.json b/jsf/tests/data/empty-list-pro.json
new file mode 100644
index 0000000..d504f1a
--- /dev/null
+++ b/jsf/tests/data/empty-list-pro.json
@@ -0,0 +1,31 @@
+{
+  "$schema": "http://json-schema.org/schema#",
+  "type": "object",
+  "properties": {
+    "content": {
+      "type": "object",
+      "properties": {
+        "list": {
+          "type": "array"
+        },
+        "non-empty-sub-list": {
+          "type": "array",
+          "items": {
+            "type": "array",
+            "items": {
+                "type": "string"
+            }
+          }
+        },
+        "sub-list": {
+          "type": "array",
+          "items": {
+            "type": "array"
+          }
+        }
+      },
+      "required": ["list", "sub-list"]
+    }
+  },
+  "required": ["content"]
+}
\ No newline at end of file
diff --git a/jsf/tests/data/empty-list.json b/jsf/tests/data/empty-list.json
new file mode 100644
index 0000000..3d97dba
--- /dev/null
+++ b/jsf/tests/data/empty-list.json
@@ -0,0 +1,8 @@
+{
+    "type": "object",
+    "properties": {
+        "items": {
+            "type": "array"
+        }
+    }
+}
\ No newline at end of file
diff --git a/jsf/tests/test_default_fake.py b/jsf/tests/test_default_fake.py
index 56b42e6..7523828 100644
--- a/jsf/tests/test_default_fake.py
+++ b/jsf/tests/test_default_fake.py
@@ -544,3 +544,43 @@ def test_use_defaults_and_examples(TestData):
         assert d["name"] in ["Chop", "Luna", "Thanos"]
         breed = d.get("breed")
         assert breed is None or breed == "Mixed Breed"
+
+def test_gen_empty_list(TestData):
+    with open(TestData / "empty-list.json") as file:
+        schema = json.load(file)
+    p = JSF(schema, allow_none_optionals=0.0)
+
+    fake_data = [p.generate(use_defaults=True, use_examples=True) for _ in range(10)]
+    for d in fake_data:
+        assert isinstance(d, dict)
+        assert "items" in d
+        assert isinstance(d["items"], list)
+        assert len(d["items"]) == 0
+
+def test_gen_empty_list_pro(TestData):
+    with open(TestData / "empty-list-pro.json") as file:
+        schema = json.load(file)
+    p = JSF(schema, allow_none_optionals=0.0)
+
+    fake_data = [p.generate(use_defaults=True, use_examples=True) for _ in range(10)]
+    for d in fake_data:
+        assert isinstance(d, dict)
+        assert "content" in d
+
+        assert isinstance(d["content"], dict)
+        assert "list" in d["content"]
+        
+        assert isinstance(d["content"]["list"], list)
+        assert len(d["content"]["list"]) == 0
+        
+        assert "non-empty-sub-list" in d["content"]
+        assert isinstance(d["content"]["non-empty-sub-list"], list)
+        assert len(d["content"]["non-empty-sub-list"]) >= 1
+        assert isinstance(d["content"]["non-empty-sub-list"][0], list)
+        assert any(len(sublist) >= 1 and all(isinstance(item, str) for item in sublist) for sublist in d["content"]["non-empty-sub-list"])
+
+        assert "sub-list" in d["content"]
+        assert isinstance(d["content"]["sub-list"], list)
+        assert len(d["content"]["sub-list"]) >= 1
+        assert isinstance(d["content"]["sub-list"][0], list)
+        assert len(d["content"]["sub-list"][0]) == 0