This study tests the hypothesis that optimum birthweight for survival is lower among hospital-born infants in Puno, Peru (altitude 3860 m) than that among their counterparts at low altitude in Tacna, Peru (altitude 600 m). The data are derived from hospital birth records for 1971 and 1972 and municipal death records for 1971 through 1973. Linking these records permits analysis of the patterns of mortality in relation to birthweight. Stabilizing selection upon birthweight is operating in both populations. The high altitude population has a lower mean birthweight and a lower optimal birthweight. The Puno population is closer to its optimal birthweight distribution and, as a result of mortality during infancy, is approaching its optimum birthweight distribution for survival more rapidly than the Tacna population. It appears that the high altitude Puno population may well be adapted to its environment in the sense that there is less selective mortality on birthweight phenotypes.