A total of 166 volunteers for an AIDS vaccine study (Vaxsyn, baculovirus produced recombinant GP160; MicroGeneSys Inc, West Haven, Connecticut, USA) were interviewed and examined. Blood was collected for routine laboratory testing as well as T-cell counts, HIV ELISA (EIA), Western blot (WB) and p24 Ag. Eighty-five men (mean age 22.2 years, range 18-42) and 81 women (mean age 23.9 years, range 17-50) volunteered; 130/166 (78%) were university students. Most had learned of the study from news media (55%), friends or workplace (37%). The most common causes for exclusion were the presence of indeterminate WB (26.5%) or a change of mind after the initial interview (24%). Other causes were abnormal cell count and differential (7.2%), elevated alanine aminotransferase (3.6%), positive hepatitis B antibody (3.6%), abnormal urinalysis (3.4%), recent venereal disease (3.0%), T4 cell count less than 400 (1.9%), abnormal chest X-ray (1.7%), recognized high-risk behaviour (1.7%), multiple sex partners (1.2%), positive rapid plasma reagin test (1.2%), failure to meet age criteria (1.2%), unable to be available for entire study (1.2%), abnormal physical examination (0.6%) and positive p24 Ag (0.6%). No volunteers had positive EIA, but 14.5% had more than one reason for exclusion. Even in a community with low prevalence for HIV, a large majority of healthy heterosexual volunteers can be expected to be ineligible for enrollment in HIV vaccine trials. An average of 4.8 volunteers were screened for each of 12 vaccinees chosen.