Opportunistic sighting records (n=14207) of 31 cetacean species complexes, species and subspecies, and 14 spatially comprehensive environmental data layers were combined to derive habitat preference models, and, for a select number of species with sufficient data, to derive relative density using a two-step hurdle model. Predicted distributions were consistent with known distributions and the respective species sighting and stranding records, although uncertainty remains for rarely sighted taxa.