In this paper, we examine the question whether criminal psychology can become an auxiliary science of criminalistics in the future. The issue is not trivial, as none of the criminalistic studies. we have examined. pays particular attention to psychology. But in the UK, criminal psychology seems already to be one of the resources of criminalistics. We examine the issue through the presentation of the first results of one of our ongoing research projects. By using a machine-learning approach we show that the motivation for homicide can be predicted with a precision of over 90% based solely on on-site clues. We believe that this, perhaps modest, research result goes beyond itself and raises the possibility that continuation and completion of the research may turn criminal psychology into a resource of criminalistics in Hungary, as well, probably the second time in the world.