Few data exist on the persistence of measles antibodies after vaccination of West African infants. Therefore we examined measles antibody titers 5 to 7 years after children in rural Senegal had received high titer Edmonston-Zagreb (EZ-HT), high titer Schwarz (SW-HT) or standard titer Schwarz (SW-STD) measles vaccines in infancy. Children had received either high titer vaccines at 5 months of age or standard titer at 10 months of age. Finger prick blood samples were tested for measles antibody 5 to 7 years later by the hemagglutinin inhibition test. Persistence of antibody after high titer vaccines was poor with the result that 39 and 50% of the EZ-HT and the SW-HT groups had low titers of hemagglutinin inhibition measles antibodies (< or =125 mIU/ml). Nineteen percent of the children in the SW-STD group had low titers which is a lower prevalence than in the high titer groups [relative risk (95% confidence intervals), 0.05 (0.28 to 0.88) vs. EZ-HT; relative risk, 0.38 (0.22 to 0.66) vs. SW-HT]. Geometric mean (95% confidence interval) antibody titers in children with detectable values were 616 (435 to 871) in the EZ-HT, 1106 (616 to 1866) in the SW-HT and 1271 (871 to 1741) mIU/ml in the SW-STD groups, respectively. Multivariant regression analysis showed that mean titers were 2.00 (1.03 to 3.89) times higher for children with low prevaccination antibody titers (< or =125 mIU/ml) and 3.06 (1.90 to 4.94) times higher if blood was collected in the rainy season. Given the rapid decline in antibody titers over a 5- to 6-year period in an area where measles vaccine coverage was high, it seems likely that multiple dose immunization schedules will be needed in the future to maintain protective antibody concentrations (>125 mIU/ml) in West Africa. The role of subclinical boosting by exposure to natural measles and the possible role of malaria, which increases immunoglobulin turnover, in influencing long term antibody persistence after vaccination deserve further investigation.