This study examined whether, and by which life course models, adolescent neighborhood environment relate to health in mid-adulthood. Data came from the Northern Swedish Cohort (n=1001), surveyed at age 16, 21, 30 and 42 years including functional somatic symptoms at age 42, and individual disadvantage neighborhood disadvantage at all four ages. Results from cross-classified multilevel models showed that 12.7% of age 42 health variance was explained by an interaction of age 16 and age 42 neighborhood of residence. Our study thus suggests that health variation by neighborhood in mid-adulthood may partly depend on neighborhood of residence in adolescence.