The Mutagroup at Jackson Labs is interested in generating new mouse models for studying neurological disease by producing mutations in mice by injecting them with ENU. The group proposes to produce large numbers of potential mutants and screen them for phenotypic anomalies. In this report we propose a statistical algorithm to flag phenotypic deviants. We have applied the algorithm to a pilot data set collected by Dr. Kevin Seburn on mice placed in cages equipped with monitoring devices. Aiming for a 5% false positive rate, the algorithm was able to detect 18 of the 27 mutant mice it was presented.