>select id1, count(distinct 'field') as cntfld form table1 group by id1

>select distinct 'field' from table1 where id1=somevalue
>
>number of records in second case is different from cntfld form first case for same value of id1

Hi Christian,

Is there any chance you've got Null values in your table?

Create Cursor TestCount (Field1 I Null, Field2 C(10) Null)
Insert Into TestCount Values (1, "One")
Insert Into TestCount Values (1, "One A")
Insert Into TestCount Values (1, .Null.)
Insert Into TestCount Values (.Null., "Null")
Insert Into TestCount Values (2, "Two")
Select Count(Distinct Field2) ;
From TestCount ;
	Where Field1 = 1    && 2

Select Distinct Field2 ;
	From TestCount ;
	Where Field1 = 1    && 3 rows