We study the relationship between environmental conditions at birth and adult stature using cohort-state level data in Brazil. We find that GDP per capita in the year of birth, not infant mortality rate, is a robust correlate of population stature in Brazil during the period 1950-1980. Our results are robust to a battery of robustness checks. Using a useful bracketing property of the (state) fixed effects and lagged dependent variables (heights) estimators, we find that an increase in GDP per capita of the magnitude corresponding to that period is associated with 43%-68% of the increase in adult height occurring in the same time span. Income, not disease, appears to be the main correlate of Brazilian population heights in the second half of the 20th Century.