Mortality was evaluated in 1972 children who had received measles vaccines at 6-11 months of age that were 10-fold (medium titer) or 100-fold (high titer) greater than standard titer. Mortality among boys did not differ by vaccine titer and was similar to mortality in children who received standard-titer vaccine. Girl recipients of high-titer vaccine had somewhat greater mortality than girls who received medium-titer vaccine (risk ratio = 1.71, 95% confidence interval = 0.91-3.24). Increased mortality was associated with high-titer vaccine for girls but not for boys (P = .04). There was no evidence of selection bias or preferential health care by sex that might explain the differential mortality. This mortality pattern has been noted in two other populations with high background infant and childhood mortality. The biologic basis for this effect on mortality has not been determined. Data from this and other studies have resulted in discontinuation of the use of high-titer measles vaccines.