Thanks for all these replies. Actually my problem was a tricky one . I figured that out. May be I did not give more clarification. The problem was , I got an user request to get data from 2 tables then create another table out of it, then he can show it in the web site.
He wants columns like productNo, description, division,transit, some other columns.., then for the quantity and date columns he wants to select by the location. location is one of the columns in the one of the 2 tables. He wants if the location is 01 then the quantity
will for that location will be put in a new column like qty_01, same for the date like date_01. The table has 6 different locations. And productNO. can have multiple locations. So when I did case like
Case table2.location when 01 then — else 0. It did not work. It will put multiple rows for the ProductNo column which he doesn’t want. So I created 3 stored procedures. One will insert into the distinct productNo, and all other columns into the new table except the qty_01,date_01 columns.Then I declare a cursor and fetch the productNo and passing that productNo I call another procedure which will get the location for that productNo then passing both productNo and location i call the third procedure which will update the qty_xx, and date_xx columns and put the values in there. This one works fine. Thanks..

Also in a multiple quantity@location scenario, the design (or data load procedure) is probably flawed to allow multiple rows with the same product number, which should be the primary key.

select ProcuctNo,Description,Sum(isnull(loc_qty_x)) as ‘loc_qty_x’,
Sum(isnull(loc_qty_y)) as ‘loc_qty_y’,Sum(isnull(loc_qty_z)) as ‘loc_qty_z’
group by ProductNo,Description where productNo = ‘blah’