>This is what I have been doing:
>
>SELECT regtemp1.username,regtemp1.companynam, regtemp1.companyid, regtemp1.settdate,;
> regtemp1.effdate, regtemp1.amount, regtemp1.batchnum, regtemp1.originid,;
> regtemp1.trancode, regtemp1.entrydescr,;
> regtemp1.reasoncode, regtemp1.orgtracnum, regtemp1.rettracnum, regtemp1.addeninfo,;
> regtemp1.bankname, regtemp1.portfolio, regtemp1.DESCRIP,;
> regtemp1.dbaname, merchantinfo.dbaname AS newdbaname,;
> regtemp1.mid, merchantinfo.mid AS newmid,;
> regtemp1.acctnumber, merchantinfo.acctnumber AS newacctnum,;
> regtemp1.tranrout, merchantinfo.tranrout AS newtranrt;
> FROM regtemp1 LEFT OUTER JOIN returnitems!merchantinfo ;
> ON ALLTRIM(regtemp1.mid) == ALLTRIM(merchantinfo.mid);
> HAVING LEFT(reasoncode,1)=="R" AND !BETWEEN(amount,nAmount1,nAmount2) AND entrydescr!="PRENOTE";
> ORDER BY regtemp1.reasoncode, regtemp1.settdate, regtemp1.effdate, regtemp1.amount,;
> regtemp1.mid;
> INTO TABLE regtemp2
>
>
>This is great if the MID matches but does me no good if say one digit is missing on the right of the MID string. I thought aout doing a thing where it adds 0-9 on the end of the exisiting MID and checking in between each add. This of course runs into the problem of 1111113 being the correct number and me adding a "2" thinking it's the right number. If I did something like this I would then have to have another way of checking the origianal data with what I find in the data warehouse records.

It would be helpful to see the length of your MID fields and a few examples of MIDs that are not matching but you wish would match.