Two-hundred and seventy lepromatous patients who had completed treatment received multiple treatments with heat-killed M. leprae and BCG and were monitored for changes in humoral responses to M. leprae-specific antigens. These patients were divided into four treatment groups: placebo (n = 69); BCG (n = 68); M. leprae only (n = 71); and BCG + M. leprae (n = 62). They were monitored for 15 months, receiving five inoculations for each treatment regimen. Two ELISA systems, one measuring antibodies to M. leprae-specific epitopes of the phenolic glycolipid I (NDO-ELISA) and the other of 36-kD protein antigens (INH-ELISA) were used to measure serological changes during this period of immunotherapy. We found no significant increase in serological reactivity with the different treatments, as measured by NDO-ELISA. INH-ELISA similarly showed no significant changes, with the exception of increased values in a small group 13% (36/270) which became skin test-positive during the course of the study. The NDO-ELISA results indicate that use of heat-killed M. leprae or BCG + heat-killed M. leprae did not stimulate the humoral response to the semi-synthetic PG-I antigens of M. leprae. Thus, the NDO-ELISA may be useful in monitoring the outcome of vaccine trials in which killed M. leprae or M. leprae fractions are used, since seroconversion may indicate disease, rather than a response to the vaccine material.