Objectives:: This study used data gathered during evaluations conducted by the Comprehensive Assessment and Training Services (CATS) Project to determine the relative contribution of four primary domains (demographic, adult characteristics, child characteristics, relational characteristics) to variation in the severity of child maltreatment, and to facilitate the development of empirically-based risk assessment tools. Methods:: A multidimensional, multi-method approach to data collection was used to assess 208 maltreating parents. Hierarchical multiple regression analyses were performed by entering each domain of variables into the regression equation in four consecutive blocks. Results:: The results of this study suggest that maltreatment severity varies as a function of the level of trauma recovery, substance misuse, a child's externalizing behavior, and qualitative features of the parent-child relationship, as well as the level of family stress. Relationship variables added to the amount of variance accounted for in the overall model, beyond demographic, adult, and child variables. Conclusions:: The findings of this study call attention to the importance of developing a multidimensional assessment model of the factors leading to child maltreatment severity and to broadening the parameters of assessment models used in child maltreatment research. The role that each of the contributing factors plays in the interference of effective parenting is also discussed.