The research goal is to analyze the determinants of dual method use among adolescents in order to better design and target prevention strategies related to STD/HIV and pregnancy. The analysis will utilize a nationally representative sample of adolescents that will permit identification of racial/ethnic variation in the determinants of dual method use. Extant research documents wide variation in the proportion of adolescents who use dual methods; yet the published literature has predominantly focused on demographic correlates analyzed atheoretically or with theories or data sets lacking contextual determinants of behavior. Using the Theory of Triadic Influences, an integrating framework for understanding the interplay among social, attitudinal, and intrapersonal influences on health behavior, analyses will explicate determinants of dual method use. Analyses will utilize one year longitudinal data from the National Longitudinal Study of Adolescent Health, including adolescent in-home interview, parent questionnaire, friendship nomination and school administrator datasets. This includes a multi-ethnic sample of 12,000 adolescents, grades 7- 12, with over-sampling of minority and special populations bringing the total sample to 20,530. The target sample will include adolescents who reported ever having sexual intercourse (N=9,427) at Time 1 or 2. The following aims guide the research plan for multivariate, longitudinal analyses: AIM I Describe patterns of dual contraceptive use within sexual relationships, over time, and across partners. Based on these patterns, empirically derive approaches to operationalizing "dual method use" of STD/HIV and pregnancy protection as a categorical as well as a continuous variable that will be used in subsequent analyses. AIM II Examine determinants of consistency of dual method use through the testing of models mapping relationships between T1 variables, including T1 dual method use, and T2 dependent variables. This includes examination of direct, moderating, and mediating effects on T2 dual method use as defined by the conceptual model. AIM III Examine T1 and T2 determinants of T2 patterns of dual method use, comparing models across gender groups and racial/ethnic groups: Euro-American, African-American, and Hispanic. The analytic plan moves from construct validation to structural equation modeling to understand mechanisms by which social, attitudinal, and intrapersonal influences affect dual method use, over time and across sexual partners. It will also contribute to the, as yet, underdeveloped knowledge base of adolescent dual method use.