diff --git a/dbt_dry_run/execution.py b/dbt_dry_run/execution.py
index 1d2085f..b76dd54 100644
--- a/dbt_dry_run/execution.py
+++ b/dbt_dry_run/execution.py
@@ -24,6 +24,7 @@
 
 def should_check_columns(node: Node) -> bool:
     check_column = node.get_combined_metadata("dry_run.check_columns")
+
     if check_column is not None:
         return bool(check_column)
 
diff --git a/dbt_dry_run/node_runner/incremental_runner.py b/dbt_dry_run/node_runner/incremental_runner.py
index 0a090e4..e864ddb 100644
--- a/dbt_dry_run/node_runner/incremental_runner.py
+++ b/dbt_dry_run/node_runner/incremental_runner.py
@@ -108,7 +108,7 @@ def get_merge_sql(
                 USING (
                   {select_statement}
                 )
-                ON True
+                ON False
                 WHEN NOT MATCHED THEN 
                 INSERT ({values_csv}) 
                 VALUES ({values_csv})
diff --git a/integration/conftest.py b/integration/conftest.py
index 11bb7e9..5b7cc67 100644
--- a/integration/conftest.py
+++ b/integration/conftest.py
@@ -37,15 +37,26 @@ def __init__(
 
     @contextmanager
     def create_state(
-        self, node: Node, columns: Iterable[str]
+        self,
+        node: Node,
+        columns: Iterable[str],
+        partition_by: Optional[str] = None,
+        require_partition_by: bool = False,
     ) -> Generator[None, None, None]:
         node_name = node.to_table_ref_literal()
         schema_csv = ",\n".join(columns)
+        partition_by_clause = f"""
+        PARTITION BY {partition_by}
+        OPTIONS (require_partition_filter = {require_partition_by})
+        """
+        if not partition_by:
+            partition_by_clause = ""
         create_ddl = f"""
-            CREATE OR REPLACE TABLE {node_name}
+        CREATE OR REPLACE TABLE {node_name}
             (
                 {schema_csv}
-            );
+            )
+            {partition_by_clause};
         """
         client: Client = self._project.get_connection().handle
         client.query(create_ddl)
diff --git a/integration/projects/test_incremental/models/required_partition_filter.sql b/integration/projects/test_incremental/models/required_partition_filter.sql
new file mode 100644
index 0000000..565ce94
--- /dev/null
+++ b/integration/projects/test_incremental/models/required_partition_filter.sql
@@ -0,0 +1,16 @@
+{{
+    config(
+        materialized="incremental",
+        on_schema_change="append_new_columns",
+        require_partition_filter = true,
+        partition_by={'field': 'snapshot_date', 'data_type': 'date'}
+    )
+}}
+
+SELECT
+   col_1,
+   col_2,
+   snapshot_date
+FROM (SELECT "foo" as col_1, "bar" as col_2, DATE("2023-01-04") as snapshot_date)
+
+
diff --git a/integration/projects/test_incremental/test_incremental.py b/integration/projects/test_incremental/test_incremental.py
index 1836bd6..b35ee33 100644
--- a/integration/projects/test_incremental/test_incremental.py
+++ b/integration/projects/test_incremental/test_incremental.py
@@ -174,3 +174,21 @@ def test_column_order_preserved_on_schema_change_append_new_columns(
             node_id,
         )
         assert_report_node_has_columns_in_order(report_node, ["col_2", "col_1"])
+
+
+def test_required_partition_filter(
+    compiled_project: ProjectContext,
+):
+    node_id = "model.test_incremental.required_partition_filter"
+    manifest_node = compiled_project.manifest.nodes[node_id]
+    columns = ["col_1 STRING", "col_2 STRING", "snapshot_date DATE"]
+    with compiled_project.create_state(manifest_node, columns, "snapshot_date", True):
+        run_result = compiled_project.dry_run()
+        assert_report_produced(run_result)
+        report_node = get_report_node_by_id(
+            run_result.report,
+            node_id,
+        )
+        assert_report_node_has_columns_in_order(
+            report_node, ["col_1", "col_2", "snapshot_date"]
+        )