diff --git a/requirements.txt b/requirements.txt
index 280274771..cbdf05784 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -8,7 +8,7 @@
 ./soda/spark[hive]
 ./soda/spark[odbc]
 ./soda/spark[databricks]
-./soda/spark_df
+./soda/spark_df[pyspark]
 ./soda/scientific[simulator]
 ./soda/sqlserver
 ./soda/mysql
diff --git a/soda/spark_df/setup.py b/soda/spark_df/setup.py
index 9c60d5139..6d4305e2b 100644
--- a/soda/spark_df/setup.py
+++ b/soda/spark_df/setup.py
@@ -8,12 +8,14 @@
 
 requires = [
     f"soda-core-spark=={package_version}",
-    "pyspark>=3.4.0",
 ]
+extras = {"pyspark": ["pyspark>=3.4.0"]}
+
 # TODO Fix the params
 setup(
     name=package_name,
     version=package_version,
     install_requires=requires,
+    extras_require=extras,
     packages=find_namespace_packages(include=["soda*"]),
 )
diff --git a/soda/spark_df/soda/data_sources/spark_df_connection.py b/soda/spark_df/soda/data_sources/spark_df_connection.py
index 752fe0a1e..2811a11a4 100644
--- a/soda/spark_df/soda/data_sources/spark_df_connection.py
+++ b/soda/spark_df/soda/data_sources/spark_df_connection.py
@@ -1,9 +1,13 @@
-from pyspark.sql.session import SparkSession
+from typing import TYPE_CHECKING
+
 from soda.data_sources.spark_df_cursor import SparkDfCursor
 
+if TYPE_CHECKING:
+    from pyspark.sql.session import SparkSession
+
 
 class SparkDfConnection:
-    def __init__(self, spark_session: SparkSession):
+    def __init__(self, spark_session: "SparkSession"):
         self.spark_session = spark_session
 
     def cursor(self) -> SparkDfCursor:
diff --git a/soda/spark_df/soda/data_sources/spark_df_contract_data_source.py b/soda/spark_df/soda/data_sources/spark_df_contract_data_source.py
index c4322fd8b..57d4fc0a4 100644
--- a/soda/spark_df/soda/data_sources/spark_df_contract_data_source.py
+++ b/soda/spark_df/soda/data_sources/spark_df_contract_data_source.py
@@ -2,8 +2,8 @@
 
 import logging
 import re
+from typing import TYPE_CHECKING
 
-from pyspark.sql import SparkSession
 from soda.data_sources.spark_df_connection import SparkDfConnection
 from soda.execution.data_type import DataType
 
@@ -13,9 +13,11 @@
 
 logger = logging.getLogger(__name__)
 
+if TYPE_CHECKING:
+    from pyspark.sql import SparkSession
 
-class SparkDfSqlDialect(SqlDialect):
 
+class SparkDfSqlDialect(SqlDialect):
     def __init__(self):
         super().__init__()
 
@@ -87,7 +89,6 @@ def regex_replace_flags(self) -> str:
 
 
 class SparkDfContractDataSource(FileClContractDataSource):
-
     def __init__(self, data_source_yaml_file: YamlFile, spark_session: SparkSession):
         data_source_yaml_dict: dict = data_source_yaml_file.get_dict()
         data_source_yaml_dict[self._KEY_TYPE] = "spark_df"
diff --git a/soda/spark_df/soda/data_sources/spark_df_cursor.py b/soda/spark_df/soda/data_sources/spark_df_cursor.py
index a73a8e67a..d16997561 100644
--- a/soda/spark_df/soda/data_sources/spark_df_cursor.py
+++ b/soda/spark_df/soda/data_sources/spark_df_cursor.py
@@ -1,7 +1,10 @@
 from __future__ import annotations
 
-from pyspark.sql import DataFrame, SparkSession
-from pyspark.sql.types import Row
+from typing import TYPE_CHECKING
+
+if TYPE_CHECKING:
+    from pyspark.sql import DataFrame, SparkSession
+    from pyspark.sql.types import Row
 
 
 class SparkDfCursor: