Probablistic matching is a method that builds the linkages among the records in two data sets which do not share any common record identifications. This presentation will focus on our research on this methodology. Some simulation results & applications to the NYSDOH SPARCS data sets will also be presented.