Although an extensive body of literature exists on the impact of Head Start, very few studies have used an experimental design with random assignment, a key methodological component needed to increase the weight of evaluation findings. This study used a randomized design with a wide range of outcomes related to school readiness to assess the impact of the Head Start program. Participating in the study were all eligible 4-year-old applicants and their parents within selected centers in a Head Start program in a southern urban setting. Participants were assigned randomly to Head Start or control group conditions after eligibility determinations had been made. Head Start participants attended the program for at least 7 months. Child assessments were made in the areas of health, social skills, cognitive skills, and preliteracy skills. Analysis techniques included growth curve modeling and traditional analysis of variance. The initial status of the two groups was equivalent; the growth rates for the Head Start children were statistically faster than for the control children for receptive vocabulary and phonemic awareness measures. There were no significant differences between the groups or change over time in social skills or positive approach to learning. For Head Start children, there were no differences in ratings of problem behaviors between the fall and the following spring; ratings of problem behaviors significantly declined for the control group during the same time period. A greater proportion of Head Start parents than control parents reported improvement in some of their child's and their own health habits. (Contains 30 references.) (KB)