Antibody (IgG) responses to mycobacterial (BCG; PPD; Mycobacterium leprae soluble antigen, MLSA) and leishmanial (Leishmania mexicana LV4) antigens were measured in 208 initially PPD and leishmanin skin-test negative volunteers divided into four vaccine groups as follows: 68 received BCG plus killed promastigotes (group A), 47 received BCG alone (group B), 47 received killed promastigotes alone (group C), and 46 formed the diluent control (placebo, group D). Three vaccine doses were administered at 8-12 week intervals. Vaccinees were bled immediately prior to each vaccination, and again at 3- and 12-month follow-up. Skin tests were performed prevaccination, and again at the 3- and 12-month follow-up. Anti-BCG, anti-PPD and anti-MLSA IgG levels increased significantly in groups A and B receiving BCG. The presence of leishmanial antigen (with BCG) in the inoculum suppressed the IgG response to Mycobacterium tuberculosis/Mycobacterium bovis-related (PPD and BCG), but not M. leprae-related (MLSA)-related, antigens. A small but significant increase (relative to prevaccination level) in response to MLSA, but not to BCG or PPD was observed in the non-BCG-vaccinated groups. The background level of response to mycobacterial and leishmanial antigens was higher in the Venezuelan vaccinees than in non-endemic (British) volunteers. Responses to leishmanial antigen were not enhanced in the two vaccine groups receiving killed promastigotes (with/without BCG) compared with the BCG alone and placebo groups. Instead, all vaccine groups showed a pattern of response consistent with either (i) a response to the skin-test antigen or, more likely, (ii) seasonal endemic exposure to leishmanial antigen. Interestingly, this endemic response to leishmanial antigen was enhanced in the vaccine groups receiving BCG, despite the fact that group B received no leishmanial antigen in the vaccine inoculum. When prevaccination IgG levels (mean + 3 standard deviations) were used to determine a negative cut-off, a low percentage (< 38%) of vaccinees converted to responder status for either anti-mycobacterial or anti-leishmanial responses, and those who did would be classified as 'low-responder' status compared with titres observed in severe forms of disease. Hence, although there was evidence for a background endemic response to both leishmanial and mycobacterial antigens, there was no evidence that vaccination per se led to a potentially disease exacerbatory level of TH2-associated antibody response especially with respect to the anti-leishmanial response.(ABSTRACT TRUNCATED AT 400 WORDS)