Another thing to watch for if selecting into a cursor is if some of your source field names are greater than 10 chars long:

1) the cursor's field will be truncated to 10 chars
2) if there are 2 fields with the same 1st 10 chars you'll get an error

>Pretty much the only times I would use a table instead of a cursor is if:
>
>1) You need to keep the results of this query after the program closes (or datasession is closed)
>
>or
>
>2) Since "Select [fields] from [tables] into cursor [cursorName]" most often results in a read only set of records, I would use "into table [tableName]" instead if I intend to modify the data in any way post-select.
>
>In my experience, I'm just using cursors a VERY high percentage of the time rather than temporary tables.