A theory-based multicomponent intervention (Gimme 5) was designed and implemented to impact fourth- and fifth-grade children's fruit, juice, and vegetable (FJV) consumption and related psychosocial variables. Gimme 5 was a randomized controlled intervention trial with school (n = 16 elementary) as unit of random assignment and analysis. Participants included the cohort of students who were in the third grade in the winter of 1994 and students who joined them in the fourth and fifth grades. The intervention included a curriculum, newsletters, videotapes, and point-of-purchase education. Evaluation included 7-day food records and psychosocial measures from students, telephone interviews with parents, and observational assessments. Favorable results were observed for consumption of FJV combined, FJV consumed at weekday lunch, eating FJV self-efficacy, social norms, asking behaviors, and knowledge. A theory-based school nutrition education program can help change children's FJV consumption and impact factors at home that predispose to FJV consumption, but changes were small, and their persistence is unknown.