>>>>Unfortunately, this would not give us only error cases - I just thought of the same SQL. You don't need to join here. This would give all C3 cases even when a/b combination is the same - I don't see a simple way to separate them - need to think.
>>>
>>>Yes, and start reading the question first
>>
>>I believe we wanted to get all records where we have discrepancy. BTW, I figured this out while driving.
>>

>>select T.* from myTable T INNER JOIN (select a,b,c from myTable group by c, a, b) X on T.c= X.c and (T.a<>X.a OR T.b <>X.b)

>
>Same defect as message#1416403
>
>try with this
>

>	
>		create cursor Schnaps ;
>	(	c1	c(1), ;
>		c2	c(1),; 
>		c3	c(1) ;
>	)
>	
>	insert into Schnaps values ('A', 'a', '1')
>	insert into Schnaps values ('A', 'b', '2')
>	insert into Schnaps values ('A', 'b', '3')
>	insert into Schnaps values ('B', 'c', '4')
>	insert into Schnaps values ('B', 'd', '4')
>	insert into Schnaps values ('B', 'e', '4') && added
>	insert into Schnaps values ('B', 'f', '4') && added
>	insert into Schnaps values ('A', 'b', '3')
>

Unfortunately, the problem is too complex - I spent ~20+ minutes and could not solve it - need real concentration - may try tonight at home.

	create cursor Schnaps ;
	(	c1	c(1), ;
		c2	c(1),; 
		c3	c(1) ;
	)
	
	insert into Schnaps values ('A', 'a', '1')
	insert into Schnaps values ('A', 'b', '2')
	insert into Schnaps values ('A', 'b', '3')
	insert into Schnaps values ('B', 'c', '4')
	insert into Schnaps values ('B', 'c', '4')
	insert into Schnaps values ('B', 'd', '4')
	insert into Schnaps values ('B', 'e', '4') && added
	insert into Schnaps values ('B', 'e', '4') && added
	insert into Schnaps values ('B', 'f', '4') && added
	insert into Schnaps values ('A', 'b', '3')
	
	select T.* from Schnaps T INNER JOIN ;
	(SELECT c3, MIN(c1) as c1_min,MAX(c1) as c1_max, ;
	MIN(c2) as c2_min, MAX(c2) as c2_max FROM Schnaps X group by 1 ;
	HAVING COUNT(distinct(c1+c2)) > 1) X ON T.c3= X.c3

I only want to select the non-matching values, such as b,d,4 and b,f,4.

BTW, in SQL Server 2005 and up this problem can be easily solved with row_number() function.