This study examines the effects of early life socioeconomic and residential conditions on adult mortality. The family and residential details of children living in rural areas of Quebec, Canada, in 1901 were linked to their subsequent ages at death using a database compiling information from the 1901 Canadian Census and Quebec vital statistics registers. Survival analysis results suggest that males raised on a farm and in a household owned by their father had lower mortality after the age of fifty than other males from rural areas. Chances for survival at older ages were not equal, however, among males whose father was a farmer. Most notably, males raised on a larger farmstead, an indicator of a higher socioeconomic status, experienced lower risk of mortality than those raised by farmers owning fewer acres. Results were widely different for females, who did not gain an advantage from being raised on a farm, wealthy or not, regardless of homeownership, but instead from having a literate father. Accounting for selection bias and shared frailty among brothers served to enhance the significance and effect size of acreage wealth and of other early life factors in the prediction of male adult mortality risk. This study provides evidence that early life effects on later life health and mortality could often be underestimated, due to a failure to account for selection and unobserved heterogeneity.