The goals of the present study were to screen for behavior problems in preschool children enrolled in Head Start and to examine the feasibility of Parent-Child Interaction Therapy (PCIT) in a randomized, controlled trial of Head Start families of children at risk for the development of later antisocial behavior. Six hundred ninety-six children, ages 3 through 5, were screened for conduct problems. Results showed that on average, children were within the nonclinical range of conduct problem behaviors, with 15% of the children scoring within the clinical range on frequency of conduct problems. Twenty-six families of children who scored within the clinical range were randomized to receive either standard Head Start care or standard care plus PCIT. Results of our study showed that in families who received PCIT plus standard Head Start care, compared to families who received standard care only, parents effectively changed their interactional style with their children, and as a consequence, children's behavior changed dramatically from outside normal limits to within normal limits. Normative data for Head Start families on several measures collected during the pretreatment screening were reported to provide a relevant source of reference for examining PCIT outcome in this population. This study demonstrated the feasibility of PCIT for treating child behavior problems in Head Start families and its potential for buffering some of the adverse effects of poverty by developing parents' confidence in their parenting. (author abstract)