A total of 101 mothers and 70 fathers with conduct problem children completed a 10-week parent-training program. Treatment effectiveness was assessed at 1 month and 1 year posttreatment based on three types of outcome variables: home observations of parent and child behaviors, parent reports of their children's behavior and teacher reports of the children's adjustment. Multiple regression analyses were performed using four predictors: parent depression, marital status or marital adjustment, socioeconomic status, and amount of negative life experiences. Results indicated that depression and the amount of negative life stress made a significant contribution to the prediction of mother and father reports of child adjustment regardless of the time of the posttreatment assessment. The combination of socioeconomic status and marital status (single versus married) made a significant contribution to mothers' behaviors with their children at both posttreatment assessments. For fathers, marital adjustment made the greatest contribution to the prediction of their critical behaviors with their children immediately at posttreatment but by 1-year follow-up, socioeconomic status emerged as a more significant predictor. For children, the best predictor of the amount of observed child deviance on the home observations was single-parent status or marital adjustment. For families who had a father present, the amount of negative life stress experienced by the family in the year since treatment was completed, was the best predictor of child deviance.