BACKGROUND: Birthweight has, in several studies, been associated with mortality in adult age, even after adjustment for available socioeconomic factors. This association has been explained as a biological result of fetal undernutrition (fetal programming), by genetic predisposition, as a result of confounding by factors related to social position and lifestyle, or by a combination of these mechanisms. This study examines the relationship between birth dimensions and all-cause and cause-specific mortality in early adulthood, taking parental lifespan and social position at time of birth into account. Furthermore, the relationship between offspring birth dimensions and parental mortality is addressed. METHODS: In all 10 753 male singletons born in the metropolitan area of Copenhagen, Denmark in 1953 whose birth certificates had been traced manually in 1965 were followed from 1968 to 2002 for information on parents' and own vital status by linkage with the Civil Registration System. Causes of death for the cohort members were identified by record linkage with the Cause-of-Death Registry. Hazard ratios and 95% CI were calculated using Cox regression models. RESULTS: Low birthweight and especially short birth length were strongly associated with adult mortality risk, but no relationship between low ponderal index at birth and mortality was found. The relationship between birth size and early adult mortality was only slightly attenuated after adjustment for early-life social position and/or maternal and paternal lifespan. The associations between birthweight/birth length and mortality were stronger for the age group 35-49 years compared with the age group 15-34 years. Cause-specific mortality was inversely related to small birth dimensions for all causes of death, but strongest and graded for death from liver cirrhosis. Offspring birth dimensions showed an inverse association with parental mortality, which was most pronounced for maternal mortality. CONCLUSIONS: The strong inverse association between birth dimensions and adult mortality, but lack of association between ponderal index and mortality, indicates that more complicated mechanisms than a general long-term detrimental effect of intrauterine growth retardation explain the association between birth size and adult mortality. The heterogeneous associations between birth dimensions and mortality in young and later adulthood, the minor differences in relative risk of cause-specific mortality, and the heterogeneity in the association between birth dimensions and maternal and paternal mortality, respectively, indicate that several mechanisms (factors related to social position, common genetic factors, and specific organ programming) may account for the relation between birth measures and later mortality.