Abstract: We have developed software for the Netherlands Forensic Institute (NFI) to improve screening and matching routine for victimidentification based on DNA profiles of the victims and the family members of reported missing persons. (So direct identifications are impossible, since no DNA of the reported missing person is available). The software, called Bonaparte, uses Bayesian networks for the analysis (i.e computes probabilities that a certain victim is actually a certain reported missing person). It is designed for effective handling of the identification process in case of a large disaster with many victims and can be applied in the missing person program. In this paper we will describe the Bayesian network approach and we discuss some of the additional features to handle events with many victims.