diff --git a/quasardb/pandas/__init__.py b/quasardb/pandas/__init__.py
index 031e2d42..224d0a9f 100644
--- a/quasardb/pandas/__init__.py
+++ b/quasardb/pandas/__init__.py
@@ -349,7 +349,17 @@ def read_dataframe(conn: quasardb.Cluster, table, **kwargs):
     # dataframes.
     dfs = stream_dataframe(conn, table, **kwargs)
 
-    return pd.concat(dfs)
+    # if result of stream_dataframe is empty this could result in ValueError on pd.concat()
+    # as stream_dataframe is a generator there is no easy way to check for this condition without evaluation
+    # the most simple way is to catch the ValueError and return an empty DataFrame
+    try:
+        return pd.concat(dfs, copy=False)
+    except ValueError as e:
+        logger.error(
+            "Error while concatenating dataframes. This can happen if result set is empty. Returning empty dataframe. Error: %s",
+            e,
+        )
+        return pd.DataFrame()
 
 
 def _extract_columns(df, cinfos):
diff --git a/tests/test_pandas.py b/tests/test_pandas.py
index f9a2b0d2..25c505f6 100644
--- a/tests/test_pandas.py
+++ b/tests/test_pandas.py
@@ -665,3 +665,22 @@ def test_retries(
             )
 
         assert seen == True
+
+
+def test_read_dataframe_empty_table_sc16881(qdbd_connection, table_name):
+    """
+    Ensures qdbpd.read_dataframe returns an empty DataFrame when the table exists but contains no data.
+
+    Previously this raised ValueError due to pd.concat() on an empty result set.
+    """
+    table = qdbd_connection.ts(table_name)
+
+    table_config = [
+        quasardb.ColumnInfo(quasardb.ColumnType.Double, "d"),
+    ]
+
+    table.create(table_config)
+
+    df = qdbpd.read_dataframe(qdbd_connection, table)
+
+    assert df.empty