The two propositions critiqued are as follows: (1) the aim of comparative research is the construction of cross-cultural generalizations about social behavior; and (2) generalizations about social behavior can and must be constructed without reference to the values, motives, and intentions of those agents whose behavior is the raw material of study. The argument that in combination these two propositions are inappropriate for cross-cultural research proceeds in four stages. First, the essay defends the claim that comparative criminology must confront the existence of cultural variation in actions and meaning. In advancing this claim, attention is given to the alleged merits of cultural relativism, the several forms of which are condensed into methodological relativism and epistemological relativism. Methodological relativism is argued to be a strategy that putatively allows an observer to generalize about criminal behavior while remaining sensitive to cultural diversity. It is then argued that this strategy occasions difficulties because it avoids the role of values in human action, while epistemological relativism involves the more serious proposition that cultural and subjective values are impossible objects for comparative generalization. Finally, the argument admits to the virtues of epistemological relativism without succumbing to some of its more pernicious consequences. To accomplish this, it is maintained that comparative criminology must transform its traditional object of study. Sixty-six notes are provided.