This digest discusses the findings of empirical studies on teacher-directed and child-initiated preschool programs. Three long-term preschool curriculum comparison studies--the High/Scope Preschool Curriculum Comparison, the Louisville Head Start Study, and the University of Illinois Study--were started in the 1970s. These studies indicated that academic outcomes for Direct Instruction programs were higher than those for child-initiated programs in the short term, but that child-initiated programs showed favorable adult outcomes. The national evaluation of Planned Variation Head Start (1969-72) included some 6,000 children at 37 sites, and its models included the Direct Instruction model and at least 2 child-initiated-activities models--the High/Scope model and the Enabler model. Findings from these studies indicated that teacher-directed groups had the highest academic achievement scores at the end of the preschool program, but the High/Scope group had the greatest IQ gains. The Follow Through Project (1967-95) was designed to follow through on Head Start by providing similar services from kindergarten through third grade. In this study, Direct Instruction students scored higher on academic achievement and other measures--a result that may be attributed to the grade level involved. Six early childhood curriculum comparison studies have been conducted in the past decade: one study contrasting High/Scope classes with non-High/Scope classes, and five studies contrasting developmentally appropriate practice emphasizing child-initiated activities and developmentally inappropriate practice emphasizing teacher-directed lessons. The relevant evidence from these studies suggests that preschool programs based on child-initiated learning activities contribute to children's short- and long-term academic and social development, while preschool programs based on teacher-directed lessons obtain a short-term advantage in children's academic development by sacrificing a long-term