Try

SELECT t1.MedRecNo AS MedRecNo1, t2.MedRecNo AS MedRecNo2,
		t1.PatientID AS PatientID1, t2.PatientID AS PatientID2
	FROM mytable t1 
		JOIN mytable t2 ON CAST(t1.MedRecNo AS int) = CAST(t2.MedRecNo AS int)
			AND t1.MedRecNo <> t2.MedRecNo

It maybe slow if table is big.

>When we installed our application, we received a data dump of patient records, this was pending our live interface feed from their patient admission system. When we got the dump, each patient has a unique code (medrecNo). When we got the dump, there were no leading zero's. Now that we have a live feed of data, it sends the medrecno with leading zero's. So after the live feed has been running for nearly two weeks, I have a bunch of duplicates, the only difference is that one has 1 or more leading zero's. I can't just delete the ones without a leading zero, because they may come into the facility tomorrow, before the live feed sends their data and they have to be available from a lookup. The medrecno is all numerics (but stored in a varchar). I was wondering if there is a way to right a query that would return me a record set that would be something like:
>
>

>MedRecNo1  MedRecNo2   PatientID1   PatientID2
>01234567   1234567     1000         15001
>

>
>Where the patientID is the PK on the patient table and the int value of the medrecnos matched
>
>
>Thanks for any help
>
>Kirk