PIP: The two major foci of HIV-1 infection in Asia, Thailand and India, have separate HIV-1 epidemics related to distinct HIV-subtypes. By the late 1980s or early 1990s, there were two distinct HIV-1 epidemics in India. One epidemic, particularly in Manipur region, was typical of intravenous (IV) drug-associated epidemics and was likely due to subtype B. The other has the patterns of a typical sexually transmitted disease and was likely due to subtype C. The two epidemics in Thailand follow a similar delineation to those in India. Yet, unlike India, a new HIV-1 has emerged, HIV-1 E, in both epidemics. Subtypes C and E viruses from Asia are associated with efficient heterosexual transmission. These HIV-1 subtypes grow considerably better in Langerhans' cells (highly prevalent in tissues of the vagina, cervix, and penis foreskin and almost absent from the rectum) than does HIV-1 subtype B. HIV-1 B viruses appear to have lost their affinity for efficient heterosexual transmission through mutation, suggesting that HIV-1 B can be maintained in groups where transmission by vaginal sex is not needed (i.e., high density of contacts with homosexuals and/or IV drug users). A homogeneous, heterosexual population with the highest incidence of new HIV-1 infections, which is exposed to the same viral subtype or strain, is the ideal test population for designing and testing an HIV-1 vaccine. The ideal HIV-1 test vaccine is the polymerized envelope antigen gp120. The safety of live attenuated HIV vaccines is doubtful. It appears that HIV-2infection provides some protection against HIV-1 infection (cross-reactivity immunity). In conclusion, trials of HIV-1 vaccines are needed but should use HIV-1 C or E strains that spread rapidly by vaginal sex. These strains can be isolated from vaginal fluids with tropism for Langerhans' cells.