This study examined three components of parent engagement in an enriched Head Start home visiting program: intervention attendance, the working alliance between parents and home visitors, and parents’ use of program materials between sessions. The study identified those family and child characteristics that predicted the different components of parent engagement, and the study tested whether those components predicted sustained growth in children’s school readiness skills across four years, from preschool through second grade. Ninety-five low-income parents with four year-old children attending Head Start (56% white; 26% black; 20% Latino; 44% girls) were randomly assigned to receive the home visiting program. Assessments included home visitor, parent, and teacher ratings, as well as interviewer observations and direct testing of children; data analyses relied on correlations and hierarchical multiple regression equations. Results showed that baseline family characteristics, such as warm parent–child interactions, and child functioning predicted both working alliance and use of program materials, but only race/ethnicity predicted intervention attendance. The use of program materials was the strongest predictor of growth in children’s literacy skills and social adjustment at home during the intervention period itself. In contrast, working alliance emerged as the strongest predictor of growth in children’s language arts skills, attention skills, and social adjustment at school through second grade, two years after the end of the home visiting intervention. To maximize intervention effectiveness across school readiness domains over time, home visiting programs need to support multiple components of parent engagement, particularly working alliance and the use of program materials between sessions.