During a measles vaccine trial in a rural area of Senegal, antibody status was examined within 10 days of exposure for 228 previously vaccinated and 313 unvaccinated children more than 12 months old who were exposed to measles at home. Thirty-six percent of the children developed clinical measles, the clinical diagnosis being confirmed for 135 of the 137 children from whom 2 blood samples were collected. Vaccine efficacy was 90% (95% confidence interval, 83 to 94%). The hemagglutinin-inhibiting antibodies (HI) or plaque neutralizing antibodies (PN) assays were equally efficient in predicting susceptibility and protection against measles. Vaccinated children who had no detectable HI or PN antibodies at exposure had significant protection against measles compared with seronegative unvaccinated children (HI vaccine efficacy, 49% (95% confidence interval, 21 to 68%); PN vaccine efficacy, 43% (95% confidence interval, 12 to 62%)). The attack rate was high for children with a titer of 40 to 125 mIU) 67% (4 of 6) of those with a positive hemagglutinin-inhibiting antibody test and 36% (13 of 36) of those with a positive PN test developed measles. Attack rates among children with HI or PN titers above 125 mIU were 2% (6 of 295) and 3% (7 of 258), respectively. Because titers of < or = 120 mIU have been found to offer little protection in another study, this antibody level may be the best screening value for assessing susceptibility and protection against measles. However, it should be noted that many seronegative vaccinated children are protected against measles infection.