Objective: Child physical abuse (CPA) has been associated with adverse adult psychosocial outcomes, although some reports describe minimal long-term effects. The search for the explanation for heterogeneous outcomes in women with CPA has led to an examination of a range of CPA-related factors, from the severity of CPA incidents to the childhood family environment. This study compares several models for predicting adult outcomes: a multidimensional CPA severity scale, the presence or absence of CPA, family environment, and childhood stresses. Methods: The effect of CPA on adult outcomes was examined among 290 community-dwelling women raised in 2-parent families. Standardized measures and a focused interview were used to collect data, with siblings as collateral informants. Results: Comparison of a multidimensional CPA severity scale to a dichotomous measure of the presence or absence of CPA showed that the severity scale did not have greater predictive value for adult outcomes than the dichotomous measure. Childhood family environment scales considerably attenuated the predictive value of the dichotomous measure of CPA, exerting a greater mediating effect on outcomes than did childhood stresses. Conclusions: The specific characteristics of a CPA experience may be less important than the occurrence of CPA and the woman's childhood family environment for predicting long-term psychosocial outcomes. Practice implications: The presence of child physical abuse is substantial and continues to increase, but the clinical significance of abuse on adult outcomes is unclear. The findings of the current study lend credence to the idea that family stresses and resources other than CPA may be crucial in understanding long-term effects in women. Hence treatment and support for victims of CPA might benefit from clinicians' exploration of the family environment. (Contains 2 tables and 1 figure.)