The effect of measles vaccine potency was evaluated among 485 children aged 6 months, and the effect of vaccine strain was evaluated among 538 children aged 3.5 months, in Kinshasa, Zaire. Children aged 6 months were randomly assigned to receive either high-titre Edmonston-Zagreb (EZ-H), potency 5.7 log10/dose, or medium-titre EZ (EZ-M), potency 4.7 log10/dose, those aged 3.5 months were randomly assigned to receive either AIK-C, potency 5.5 log10/dose, or EZ-H, and were revaccinated with EZ-M vaccine at age 9.5 months. Measles antibodies were measured using the plaque reduction neutralization assay. Among children vaccinated at age 6 months, the seroresponse was significantly higher after EZ-H than EZ-M vaccine, with 92 and 83% seroconverting by 6 months postvaccination and 59 and 40% respectively having antibody titres > 200 mIU. Among children vaccinated at age 3.5 months, only 24% (AIK-C) and 22% (EZ-H) attained antibody titres > or = 200 mIU 6 months postvaccination. After revaccination at age 9.5 months, 81% of children in the AIK-C group and 73% in the EZ-H group had antibody levels > 200 mIU (p = 0.056). A retrospective survey was conducted in January 1993 to determine the mortality experience of vaccine groups, and information was obtained for 94% of the children. A total of 44 deaths (4%) were identified, with no significant differences between groups when stratified by age at vaccination.(ABSTRACT TRUNCATED AT 250 WORDS)