When blood samples were analyzed for seroconversion after measles vaccination, it was discovered that the vaccine had been ineffective for a certain period. During the 2 years between vaccination and the time of seroanalysis, nonseroconverters had a significantly higher mortality than seroconverters (P less than 0.05). The incidence of measles among nonseroconverters was 30% during the period. Between 9 months and 3 years of age, cumulative mortality was 15.1% for nonseroconverters and 4.5% for seroconverters. The difference in mortality was larger when high risk groups (twins, motherless children) were excluded from the analysis (P less than 0.01). The difference in mortality was particularly marked among children vaccinated in the age group 9 to 11 months. This as well as other community studies suggest that measles vaccination reduces child mortality from the age of vaccination by at least 30%.