diff --git a/doc/source/whatsnew/v2.1.0.rst b/doc/source/whatsnew/v2.1.0.rst
index f3b9b0067bb34..f8ad5708ace8b 100644
--- a/doc/source/whatsnew/v2.1.0.rst
+++ b/doc/source/whatsnew/v2.1.0.rst
@@ -401,6 +401,7 @@ I/O
 - Bug in :func:`read_hdf` not properly closing store after a ``IndexError`` is raised (:issue:`52781`)
 - Bug in :func:`read_html`, style elements were read into DataFrames (:issue:`52197`)
 - Bug in :func:`read_html`, tail texts were removed together with elements containing ``display:none`` style (:issue:`51629`)
+- Bug in :func:`read_sql` when reading multiple timezone aware columns with the same column name (:issue:`44421`)
 - Bug when writing and reading empty Stata dta files where dtype information was lost (:issue:`46240`)
 
 Period
diff --git a/pandas/io/sql.py b/pandas/io/sql.py
index ebb994f92d8ad..51cc3eacae284 100644
--- a/pandas/io/sql.py
+++ b/pandas/io/sql.py
@@ -131,13 +131,13 @@ def _parse_date_columns(data_frame, parse_dates):
     # we want to coerce datetime64_tz dtypes for now to UTC
     # we could in theory do a 'nice' conversion from a FixedOffset tz
     # GH11216
-    for col_name, df_col in data_frame.items():
+    for i, (col_name, df_col) in enumerate(data_frame.items()):
         if isinstance(df_col.dtype, DatetimeTZDtype) or col_name in parse_dates:
             try:
                 fmt = parse_dates[col_name]
             except TypeError:
                 fmt = None
-            data_frame[col_name] = _handle_date_column(df_col, format=fmt)
+            data_frame.isetitem(i, _handle_date_column(df_col, format=fmt))
 
     return data_frame
 
diff --git a/pandas/tests/io/test_sql.py b/pandas/tests/io/test_sql.py
index a864903d9c6d1..7a3f7521d4a17 100644
--- a/pandas/tests/io/test_sql.py
+++ b/pandas/tests/io/test_sql.py
@@ -2890,6 +2890,44 @@ def test_schema_support(self):
             res2 = pdsql.read_table("test_schema_other2")
             tm.assert_frame_equal(res1, res2)
 
+    def test_self_join_date_columns(self):
+        # GH 44421
+        from sqlalchemy.engine import Engine
+        from sqlalchemy.sql import text
+
+        create_table = text(
+            """
+        CREATE TABLE person
+        (
+            id serial constraint person_pkey primary key,
+            created_dt timestamp with time zone
+        );
+
+        INSERT INTO person
+            VALUES (1, '2021-01-01T00:00:00Z');
+        """
+        )
+        if isinstance(self.conn, Engine):
+            with self.conn.connect() as con:
+                with con.begin():
+                    con.execute(create_table)
+        else:
+            with self.conn.begin():
+                self.conn.execute(create_table)
+
+        sql_query = (
+            'SELECT * FROM "person" AS p1 INNER JOIN "person" AS p2 ON p1.id = p2.id;'
+        )
+        result = pd.read_sql(sql_query, self.conn)
+        expected = DataFrame(
+            [[1, Timestamp("2021", tz="UTC")] * 2], columns=["id", "created_dt"] * 2
+        )
+        tm.assert_frame_equal(result, expected)
+
+        # Cleanup
+        with sql.SQLDatabase(self.conn, need_transaction=True) as pandasSQL:
+            pandasSQL.drop_table("person")
+
 
 # -----------------------------------------------------------------------------
 # -- Test Sqlite / MySQL fallback