This paper’s central argument is that men’s domestic-violence treatment programs often fail to link program rationale, design, and delivery methods to a conceptualization of general and specific causes of violent and abusive behavior within an intimate relationship. There are models of behavioral change that can be applied to domestically violent men, and stages of change in domestic-violence perpetrators can be reliably measured. Issues of program integrity can be addressed by applying these behavioral change models to the design and delivery of intervention programs for men who perpetrate domestic violence. Still, there are distinctive features of domestic-violence behavior that must be addressed in treatment, including physical and sexual aggression as well as verbal and emotional abuse. The interpersonal dynamics of the relationship in which the abuse occurs must also be considered. In addition, the mandatory nature of many domestic-violence treatment programs may mean that participants lack the motivation or awareness of their need to change. Based on this analysis, the authors recommend the development of treatment responses derived from a detailed assessment of the nature and causes of the offenses, as well as the risk and likely severity of reoffending. One approach to differentiated treatment responses is through the classification of perpetrators into typologies; for example, four types of domestic-violence offenders--the generally violent aggressor, family-only, low-level antisocial, and borderline/dysphoric--have been identified by Holtzworth-Munrow, Meehan, Herron, Rehman, and Stuart (2000). An alternative approach that is being used in the treatment of sexual offenders is to define offender needs in terms of specific offending behavior rather than the characteristics of the offender. 58 references