I ended up going with a solution of using DBI for the CSV access and Win32::ODBC to do the lookup to the SQL server. It's not 100% efficient by any means (I'm still pretty new to perl and all), but it gets the job done. To do the compares, I read in the CSV record, look up the corresponding record on the database, and do a field-by-field compare, which works ourt pretty nicely. I actually did get the CSV to CSV lookups working, but it was horribly slow. Thanks for the responses.