Both child and adult socio-economic position (SEP) predict adult mortality, but little is known about the variation in the impact of SEP across the life course. The Uppsala Birth Cohort Study is a representative birth cohort born 1915-1929 in Uppsala, Sweden. For the 5,138 males and 5,069 females alive in 1980, SEP was available at birth; in adulthood (age 31-45); and in later life (age 51-65). Follow-up for mortality (all-cause, and circulatory disease) was from 1980 to 2002. To test which life course model best described the association between SEP and mortality, we compared the fit of a series of nested Cox proportional hazards regression models (representing either the critical, accumulation or sensitive period models) with a fully saturated model. For all-cause mortality in both genders, the sensitive period model best described the influence of SEP across the life course with a heightened effect in later adult life (males: Hazard Ratio (95 % CI) for advantaged SEP: 0.89 (0.81-0.97) at birth, 0.90 (0.81-0.98) in adulthood, 0.74 (0.67-0.82) in later life; females: 0.87 (0.78-0.98), 0.95 (0.86-1.06), 0.73 (0.64-0.83)). The effect of SEP on circulatory diseases mortality in males was cumulative (HR: 0.84 (0.80-0.87) per unit time in advantaged SEP). For circulatory disease mortality among females, a sensitive period model was selected due to SEP in later adult life (HR: 0.64 (0.52-0.80)). These findings suggest that reducing inequality throughout the life course might reduce all-cause and circulatory disease mortality.