Objective: To determine whether birth weights are lower at high altitudes, and whether gestational age at birth and a population's length of residence mitigate the effect of high altitude. Methods: The birth weights of 84,173 neonates recorded in the Peruvian Perinatal Information System Database were analyzed between 1995 and 2002 for the cities of Lima (150 m), Huancayo (3280 m), Cuzco (3400 m), and Juliaca (3800 m). Results: Birth weight was lower at high altitude, but there was no linear relation between altitude of residence and birth weight. Mean birth weight was higher in Juliaca than in Huancayo. There were no significant differences between the 4 cities regarding birth weights of infants born between 28 and 35 weeks of gestation. However, for infants born between 36 and 42 weeks, birth weight was lower at higher altitudes. This may be due to inadequate maternal oxygenation later in pregnancy at high altitude. In the multivariate analysis, after controlling for maternal age, marital status, parity, body mass index, pre-eclampsia or hemorrhage during pregnancy, and education, as well as sex of the newborn and gestational age at birth, birth weight was lower in all cities located at a higher altitude than Lima. Yet, longer residence at high altitudes may play a protective role. Juliaca (3800 m), where the population has resided the longest, had the lowest reduction in birth weight compared with Lima (150 m); Cuzco had intermediate values; and Huancayo (3280 m), where the population has resided the shortest, had the highest reduction in birth weight. Conclusions: Birth weight reduction, which is independent of socioeconomic factors, occurs only in births at term and may be less severe in populations that have resided longer at high altitudes. © 2006 International Federation of Gynecology and Obstetrics.