Characteristics of scientific research-careful definition of the problem, random sample selection, relevant control groups, careful measurement techniques, and caution in drawing conclusions from the data-are employed to evaluate the argument for constitutional, genetic, mental, endocrinal, and neurological theories of crime causation. A historical continuity in assumptions and methodological flaws is revealed between contemporary studies of the biological correlates of crime and their earlier counterparts. Social and political ramifications of biomedical research are considered.