Jim,
The having clause is for restricting the query for DUPLICATE records only.
i.e the resultant record set would include only those values of 'cfieldname' which occur MORE than ONCE.

*---Ranjit

>>SELECT cfieldname, count(*) AS dup_num ;
>> FROM ctable ;
>> GROUP BY cfieldname ;
>> HAVING count(*) > 1 ;
>> INTO CURSOR ctempfile
>Ranjit,
>
>Out of curiosity, why the HAVING clause? By definition there will always be at least one record for each value of the field being grouped on. Doesn;t the HAVING clause just slow things down without improving anything, or am I missing something here?