DELETE * from table1 where rowId not in( Select min(rowId), column1, column2 from table1 group by column1, column2 having count(1) > 1 ). If you want to classify the duplicates those rows that have the same value for these columns.