The author suggests that clinicians should be aware of the vast diversity in the abuser population and the need to individualize treatment appropriate to the client. Recidivism outcome studies are reviewed from the perspective of their contribution to decisions on what treatment appears to be effective with which populations rather than in terms of whether they support generalized conclusions about treatment efficacy. In addition, studies addressing the scope of deviance, nonsexual antisocial behavior, treatment outcome, prediction of compliance, and coerced treatment are examined. The author concludes that, for some abusers, behavior exists within an extensive network of paraphilias or a longstanding pattern of antisocial behavior or psychopathy; for other abusers, behavior is episodic or encapsulated. He also contends that abuser characteristics are influenced by referral patterns and settings which must be considered in addressing the utility and generalizability of research. 29 references and 6 notes