diff --git a/CHANGELOG.md b/CHANGELOG.md
index d910c646..ed59a8db 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -12,6 +12,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - `datacontract test --output-format junit --output TEST-datacontract.xml` Export CLI test results
   to a file, in a standard format (e.g. JUnit) to improve CI/CD experience (#650)
 - `dbt` & `dbt-sources` export formats now support the optional `--server` flag to adapt the DBT column `data_type` to specific SQL dialects
+- Duckdb Connections are now configurable, when used as Python library (#666)
 
 ### Changed
 
diff --git a/datacontract/data_contract.py b/datacontract/data_contract.py
index 1cef9150..1ebe8126 100644
--- a/datacontract/data_contract.py
+++ b/datacontract/data_contract.py
@@ -4,6 +4,8 @@
 if typing.TYPE_CHECKING:
     from pyspark.sql import SparkSession
 
+from duckdb.duckdb import DuckDBPyConnection
+
 from datacontract.breaking.breaking import (
     info_breaking_changes,
     models_breaking_changes,
@@ -39,6 +41,7 @@ def __init__(
         server: str = None,
         publish_url: str = None,
         spark: "SparkSession" = None,
+        duckdb_connection: DuckDBPyConnection = None,
         inline_definitions: bool = True,
         inline_quality: bool = True,
         ssl_verification: bool = True,
@@ -50,6 +53,7 @@ def __init__(
         self._server = server
         self._publish_url = publish_url
         self._spark = spark
+        self._duckdb_connection = duckdb_connection
         self._inline_definitions = inline_definitions
         self._inline_quality = inline_quality
         self._ssl_verification = ssl_verification
@@ -146,7 +150,7 @@ def test(self) -> Run:
                 inline_quality=self._inline_quality,
             )
 
-            execute_data_contract_test(data_contract, run, self._server, self._spark)
+            execute_data_contract_test(data_contract, run, self._server, self._spark, self._duckdb_connection)
 
         except DataContractException as e:
             run.checks.append(
diff --git a/datacontract/engines/data_contract_test.py b/datacontract/engines/data_contract_test.py
index cc59fee8..ae3b84b1 100644
--- a/datacontract/engines/data_contract_test.py
+++ b/datacontract/engines/data_contract_test.py
@@ -1,5 +1,7 @@
 import typing
 
+from duckdb.duckdb import DuckDBPyConnection
+
 from datacontract.engines.data_contract_checks import create_checks
 
 if typing.TYPE_CHECKING:
@@ -20,6 +22,7 @@ def execute_data_contract_test(
     run: Run,
     server_name: str = None,
     spark: "SparkSession" = None,
+    duckdb_connection: DuckDBPyConnection = None,
 ):
     if data_contract_specification.models is None or len(data_contract_specification.models) == 0:
         raise DataContractException(
@@ -43,7 +46,7 @@ def execute_data_contract_test(
     # TODO check server credentials are complete for nicer error messages
     if server.format == "json" and server.type != "kafka":
         check_jsonschema(run, data_contract_specification, server)
-    check_soda_execute(run, data_contract_specification, server, spark)
+    check_soda_execute(run, data_contract_specification, server, spark, duckdb_connection)
 
 
 def get_server(data_contract_specification: DataContractSpecification, server_name: str = None):
diff --git a/datacontract/engines/soda/check_soda_execute.py b/datacontract/engines/soda/check_soda_execute.py
index 09ad281b..9bc42ddd 100644
--- a/datacontract/engines/soda/check_soda_execute.py
+++ b/datacontract/engines/soda/check_soda_execute.py
@@ -1,6 +1,12 @@
 import logging
+import typing
 import uuid
 
+if typing.TYPE_CHECKING:
+    from pyspark.sql import SparkSession
+
+from duckdb.duckdb import DuckDBPyConnection
+
 from datacontract.engines.soda.connections.bigquery import to_bigquery_soda_configuration
 from datacontract.engines.soda.connections.databricks import to_databricks_soda_configuration
 from datacontract.engines.soda.connections.duckdb_connection import get_duckdb_connection
@@ -14,7 +20,13 @@
 from datacontract.model.run import Check, Log, ResultEnum, Run
 
 
-def check_soda_execute(run: Run, data_contract: DataContractSpecification, server: Server, spark):
+def check_soda_execute(
+    run: Run,
+    data_contract: DataContractSpecification,
+    server: Server,
+    spark: "SparkSession" = None,
+    duckdb_connection: DuckDBPyConnection = None,
+):
     from soda.common.config_helper import ConfigHelper
 
     ConfigHelper.get_instance().upsert_value("send_anonymous_usage_stats", False)
@@ -30,7 +42,7 @@ def check_soda_execute(run: Run, data_contract: DataContractSpecification, serve
     if server.type in ["s3", "gcs", "azure", "local"]:
         if server.format in ["json", "parquet", "csv", "delta"]:
             run.log_info(f"Configuring engine soda-core to connect to {server.type} {server.format} with duckdb")
-            con = get_duckdb_connection(data_contract, server, run)
+            con = get_duckdb_connection(data_contract, server, run, duckdb_connection)
             scan.add_duckdb_connection(duckdb_connection=con, data_source_name=server.type)
             scan.set_data_source_name(server.type)
         else:
diff --git a/datacontract/engines/soda/connections/duckdb_connection.py b/datacontract/engines/soda/connections/duckdb_connection.py
index 2790e1a8..f05fce2f 100644
--- a/datacontract/engines/soda/connections/duckdb_connection.py
+++ b/datacontract/engines/soda/connections/duckdb_connection.py
@@ -4,11 +4,21 @@
 import duckdb
 
 from datacontract.export.csv_type_converter import convert_to_duckdb_csv_type
+from datacontract.model.data_contract_specification import DataContractSpecification, Server
 from datacontract.model.run import Run
 
 
-def get_duckdb_connection(data_contract, server, run: Run):
-    con = duckdb.connect(database=":memory:")
+def get_duckdb_connection(
+    data_contract: DataContractSpecification,
+    server: Server,
+    run: Run,
+    duckdb_connection: duckdb.DuckDBPyConnection = None,
+):
+    if duckdb_connection is None:
+        con = duckdb.connect(database=":memory:")
+    else:
+        con = duckdb_connection
+
     path: str = ""
     if server.type == "local":
         path = server.path