>Hi Sergey,
> I think I understand. So are you saying that SELECT DISTINCT only eliminates rows for which all columns are the same?
>
> I thought that the statement:
>SELECT DISTINCT ColumnA ColumnB
>would return only 1 row for each unique value of ColumnA.
>
>Are you saying it would return 1 row for each unqiue value of ColumnA and ColumnB?
>
>Thanks, Chaim

Yes, it will return row for each unique combination of ColumnA and ColumnB. It's the same as doing GROUP BY ColumnA,ColumnB