Three different batches of live attenuated measles virus vaccine (Schwarz strain) containing, respectively, 1000, 5000 and 12 500 TCID 50 per dose, were given to groups of 129, 112 and 61 allegedly measles-negative children aged from one to six years. Pre- and postvaccination blood specimens for measles HI-tests were taken from all the vaccinees. High (96–97%) conversion rates were obtained in previously seronegative children immunized with either of the two more potent vaccines, while the vaccine containing 1000 TCID 50, although still acceptable for use according to official U.S.A. requirements, gave conversion only in 86%. There was no significant difference between the three vaccines in geometric mean postvaccination titres of seroconverting vaccinees. A follow-up study, performed two years after vaccination of another group of children with a vaccine containing about 2500 TCID 50 per dose, showed that 130 out of 138 investigated children had demonstrable measles HI-antibodies with titres closely similar to those found in a control group of naturally immune children. The incidence of febrile reactions and rash following vaccination was, respectively, about two and four times higher in seroconverting children vaccinated with the strongest vaccine used (12 500 TCID 50 per dose) than in children given the two other vaccines tested. A vaccine strength of about 5000 TCID 50 per dose, that was found to induce more than 95% conversions while still giving comparatively few side-reactions, would thus appear to be most suitable for routine vaccination of children against measles. However, in view of the relative temperature and storage instability of live attenuated measles virus, even lyophilized, it might be difficult in practice to provide such an exactly defined and reproducible content of viable virus in the vaccine at the time when it is actually being used. The findings stress the importance of adequate calibration of the live virus content in commercially available measles vaccines.