Longitudinal data from the Youth Development Study are used to examine (1) how young people establish work with self-identified career potential and how these patterns are linked to educational attainments; and (2) how adolescent achievement orientations, experiences in school and work, and sociodemographic background distinguish youth who establish themselves in careers and those who flounder during this transition. Multilevel latent class models reveal four school-to-work pathways from ages 18–31: two groups that attain careers through postsecondary education (via bachelor's or associate's–vocational degrees) and two groups that do not (distinguished by attempting college). Multinomial logistic regression models demonstrate that academic orientations, socioeconomic background, and steady paid work during high school help adolescents avoid subsequent floundering during the school-to-work transition.