Try

select iproductid, caccount, count(caccount) as cnt, 
		( SELECT TOP 1 COUNT(name)  
			FROM customer c2 
			WHERE c2.iproductid = c1.iproductid 
				AND c2.caccount = c2.caccount 
			GROUP BY name ORDER BY COUNT(name) DESC) AS AS namecount
	from customer c1
	group by iproductid,caccount
	having count(caccount)>2
	order by iproductid,cnt

>I have a query that finds duplicate customers using the same credit card:
>

>select iproductid,caccount,count(caccount) as cnt
>from customer c
>group by iproductid,caccount
>having count(caccount)>2
>order by iproductid,cnt
>

>
>This gives me results like
>

>
>iproductid  caccount                       cnt
>----------- ------------------------------ -----------
>1           xxx                            3
>1           yyy                            3
>
>

>
>Now I want another column that tells me "of those customers using the same credit card, how many have the same last name?"
>
>I probably could do it with cursors and maybe a UDF, but is there another way using the select? When I try I correlated subquery, I can't get the number of records to be returned to me.
>TIA