This paper examines the effects of a public-health intervention program on sex differentials in health and mortality during childhood. Among the different health-service packages offered as part of the experimental design, those including nutritional services seem to have been more successful in reducing excess female mortality. The reason for this success appears to have been careful follow-up of undernourished children by project workers. The results also indicate that, consistent with earlier research, girls with surviving older sisters had higher mortality rates after their first month of life. Contrary to earlier research, however, boys with surviving older brothers also have higher mortality rates, at least between the ages of one and three years. We conclude that these effects for boys and girls cannot be attributed to problems associated with larger family size, since the number of older siblings of the opposite sex (regardless of survival status) does not generally appear to be related to children's chances of survival.