The fixated/regressed typology of child molesters has been widely used in classifying molesters. Although the terminology is psychodynamic in origin, cognitive correlations for the two types can be identified. A research instrument designed to assess child molester cognitions, attitudes, and background characteristics was administered to 94 molesters, one-third of whom had been identified as fixated and two-thirds as regressed. Data analysis yielded results which were generally supportive of the criteria for the two types. Fixated molesters were likely to be more child-centered, to molest male children outside the family, to come from "broken" homes, and to use alcohol less frequently than regressed molesters. Regressed molesters were likely to be better adjusted, to molest female children outside the family and to come from intact homes.