Serum specimens from leprosy patients, their contacts and controls were examined for the presence of phenolic glycolipid I (PGL-I), a Mycobacterium leprae specific antigen, and antibodies to the antigen using enzyme-linked immunosorbent assays. Of 12 lepromatous patients with less than 2 years of therapy, 11(91.7%) were seropositive to PGL-l, thus indicating that new lepromatous cases can be identified by detecting anti-PGL-l antibodies. In contrast 88(56.4%) of 156 lepromatous patiens treated more than 2 years were positve. Moreover, only 69(40.8%) were seropositve among 169 lepromatous patients in the leprosy resettlement villages. The mean antibody level also declined significantly in proportion to the duration of chemotherapy. This may suggest the possibility of monitoring chemotherapy by detecting anti-PGL-l antibodies. The prevalence of anti-PGL-l antibodies among 200 controls from a high endemic area for leprosy was 5.5% and was significantly higher than that(1.5%) among 200 controls from a low endemic area. Of 103 household contacts in the resettlement villages, 10(9.7%) were seropositive, reflecting the frequent chance of exposure to M. leprae. However, PGL-l was not detected many in any of the sera from controls, contacts, and inactive lepromatous patients having the anti-PGL-l antibodies; on the other hand, 6(50%) of 12 lepromatous patients treated less than 2 years had detectable PGL-l in their sera. The results thus indicate that PGL-l detection may be more suitable for monitoring the effectiveness of chemotherapy and that it may be necessary to examine for the presence of PGL-l in sera from contacts and normal populations for confirming M. leprae infection.