diff --git a/data/data-pipeline/data_pipeline/etl/score/constants.py b/data/data-pipeline/data_pipeline/etl/score/constants.py
index ae1d70a1..29dac8f2 100644
--- a/data/data-pipeline/data_pipeline/etl/score/constants.py
+++ b/data/data-pipeline/data_pipeline/etl/score/constants.py
@@ -63,6 +63,13 @@ SCORE_DOWNLOADABLE_ZIP_FILE_PATH = (
 # Column subsets
 CENSUS_COUNTIES_COLUMNS = ["USPS", "GEOID", "NAME"]
 
+# Percent prefixes for rounding
+PERCENT_PREFIXES_SUFFIXES = [
+    "Percent",
+    "Percentage",
+    field_names.PERCENTILE_FIELD_SUFFIX,
+]
+
 TILES_ROUND_NUM_DECIMALS = 2
 # Tiles data: full field name, tile index name
 TILES_SCORE_COLUMNS = {
diff --git a/data/data-pipeline/data_pipeline/etl/score/etl_score_post.py b/data/data-pipeline/data_pipeline/etl/score/etl_score_post.py
index 30bcf493..b7168e67 100644
--- a/data/data-pipeline/data_pipeline/etl/score/etl_score_post.py
+++ b/data/data-pipeline/data_pipeline/etl/score/etl_score_post.py
@@ -240,12 +240,35 @@ class PostScoreETL(ExtractTransformLoad):
     def _create_downloadable_data(
         self, score_county_state_merged_df: pd.DataFrame
     ) -> pd.DataFrame:
-        df = score_county_state_merged_df[constants.DOWNLOADABLE_SCORE_COLUMNS]
+        df = score_county_state_merged_df[
+            constants.DOWNLOADABLE_SCORE_COLUMNS
+        ].copy()
 
-        float_columns = df.select_dtypes(include=["float64"])
+        float_columns = df.select_dtypes(include=["float64"]).columns
 
-        # score_tiles[constants.TILES_SCORE_FLOAT_COLUMNS] = score_tiles[
-        #     constants.TILES_SCORE_FLOAT_COLUMNS
+        # convert percentile_columns
+        percent_target_columns = []
+        for x in float_columns:
+            for col in constants.PERCENT_PREFIXES_SUFFIXES:
+                if col in x:
+                    percent_target_columns.append(x)
+
+        df[percent_target_columns] = df[percent_target_columns].apply(
+            func=lambda series: floor_series(
+                series=series * 100,
+                number_of_decimals=constants.TILES_ROUND_NUM_DECIMALS,
+            )
+        )
+
+        # # convert percentile_columns
+        # non_percentile_float_columns = [
+        #     x
+        #     for x in float_columns
+        #     if x not in constants.PERCENT_PREFIXES_SUFFIXES
+        # ]
+
+        # df[non_percentile_float_columns] = df[
+        #     non_percentile_float_columns
         # ].apply(
         #     func=lambda series: floor_series(
         #         series=series,
@@ -254,8 +277,6 @@ class PostScoreETL(ExtractTransformLoad):
         #     axis=0,
         # )
 
-        # [x for x in df.columns if field_names.PERCENTILE_FIELD_SUFFIX in x]
-
         return df
 
     def transform(self) -> None: