The question is: why do you want to make it to work in VFP8 if it works in 9? Or was it just a sample and you need something similar elsewhere for a different purpose?

>This is the solution that I was trying to avoid:
>
>SELECT crsTest.Fld1, SUM(crsTest1.Amount) AS Amount1, 0000000000.00 AS Amount2;
> FROM crsTest;
> INNER JOIN crsTest1 ON crsTest.Fld1 = crsTest1.Fld1;
> GROUP BY 1;
>UNION ;
> SELECT crsTest.Fld1, 0000000000.00 AS Amount1, SUM(crsTest2.Amount) AS Amount2;
> FROM crsTest;
> INNER JOIN crsTest2 ON crsTest.Fld1 = crsTest2.Fld1;
> GROUP BY 1;
>ORDER BY 1 ;
>INTO CURSOR crsResult
>
>SELECT Fld1, SUM(Amount1) AS Amount1, SUM(Amount2) AS Amount2;
>GROUP BY 1;
>FROM crsResult
>
>Thanks.
>
>>You got exactly what you ask for :-)
>>The doubled result in the result is because the recordset looks like this (Lets change the second 10 in the crsTest1 to 11):
>>

>>Fld1     Test1.Ammount   Test2.Ammount
>>---------------------------------------
>>1            10                100
>>1            11                100
>>

>>You got as many records as the largest number of records in all tables. In VFP9 you could use derived tables to got the result you want, I am not sure about VFP8 (can't test it right now).
>>Here VFP9 query:
>>

>>SELECT crsTest.Fld1, cTest1.Amnt AS Ammount1, cTest2.Amnt AS Ammount2;
>>       FROM crsTest;
>>INNER JOIN (SELECT Fld1, SUM(Ammount) AS Amnt;
>>                   FROM crsTest1;
>>                   GROUP BY Fld1) cTest1;
>>ON crsTest.Fld1 = cTest1.Fld1;
>>INNER JOIN (SELECT Fld1, SUM(Ammount) AS Amnt;
>>                   FROM crsTest2;
>>                   GROUP BY Fld1) cTest2;
>>ON crsTest.Fld1 = cTest2.Fld1
>>

>>
>>If this didn't works in VFP8, you could first query second two tables and then join the results to main one.