In this study we sought to explore the association between preschool instructional approach and health and well-being at age 35 for a large sample of low-income children. Participants included 989 low-income, minority children who attended Child-Parent Center preschools as part of the Chicago Longitudinal Study from 1983 to 1985. Preschool curriculum was obtained from teacher reports and validated by program evaluators. These data were categorized by raters as: high teacher-directed and child-initiated; low teacher-directed and child-initiated; low teacher-directed and high child-initiated; or high teacher-directed and low child-initiated. Data on adult outcomes were obtained through surveys and administrative records. Those in preschool classrooms with high teacher-directed and child-initiated instruction had increased odds of having a livable wage (Odds Ratio(OR) = 2.02, p = 0.001), and decreased odds of felony arrest (OR = 0.39; p < 0.001), jail or incarceration (OR = 0.35, p = 0.001), and conviction (OR = 0.52, p = 0.002) at age 35 than those in low teacher-directed and child-initiated classrooms. Participants experiencing low teacher-directed and high child-initiated instruction had increased odds of having a livable wage (OR = 2.01, p = 0.002) and decreased odds of felony arrest (OR = 0.46; p < 0.001), jail or incarceration (OR = 0.53; p = 0.023), and conviction (OR = 0.57, p = 0.01) at age 35. Findings were consistent across many model specifications and adjustments for potential attrition bias. Child-initiated instruction in preschool is a robust predictor of adulthood well-being. Early education prevention efforts to establish a blend of child-initiated and teacher-directed teaching philosophies affords the opportunity for long-term impacts on economic and criminal outcomes in adulthood. (author abstract)