Abstract Because early adolescent substance use can be a symptom of current and future maladjustment, the stakes of alcohol and marijuana use among young adolescents in the U.S. are high. As such, elucidating the etiology of early use serves an important public health purpose. In this spirit, this R03 project will delve into an ecological setting linked to substance use: family structures with only one biological parent. Past research documents that adolescents who live with only one biological parent (especially a single father and somewhat regardless of the presence of a stepparent) tend to have higher levels of alcohol and marijuana use than youth living with both biological parents through stress-related, family-based, and peer-based mechanisms. Drawing on life course theory, this project will innovatively expand on this foundational knowledge by focusing on especially young and vulnerable adolescents and by exploring variability in the links among family structure (single mother and single father vs. two biological parent and stepparent families), the focal mechanisms, and substance use across multiple layers context. It will do so by applying advanced statistical techniques to several sources of public data that have not yet been examined collectively. First, conceptualizing contextualization as historical time, joinpoint regression will be applied to repeated cross-sectional data on U.S. 14 year olds from Monitoring the Future. These analyses will identify non-linear increases in the associations between single parent families and young adolescents' alcohol and marijuana use from the early 1990s through today. Second, conceptualizing contextualization as developmental and family history, time-varying effects modeling will be applied to panel data on U.S. 14 year olds from the National Longitudinal Survey of Youth 1979-Children and Young Adults (NLSY-CYA). These analyses will identify increases in the focal associations as a function of the recency of transitions into such structures and the number of other family transitions experienced since birth. Third, conceptualizing contextualization as societal settings, structural equation modeling will be applied to cohort data on 14 year olds from NLSY-CYA, Growing up in Australia, Growing up in Ireland, and the U.K. Millennium Cohort. Using both a priori harmonization, fixed effects integrative data analysis, and missing data estimation to address differences across datasets, these analyses will identify potential increases in the focal associations in the U.S. compared to three other English-speaking countries with similar policy regimes but different norms of substance use, family structure patterns, and public supports for families. Conducted by a senior sociologist and junior psychologist with histories of productive federally funded research on families and adolescent health along with consultants from each of the non-U.S. countries under study, this project will ask and answer contextual q...