In order to examine the efficiency of an AIDS vaccine potentially acceptable for human use we have investigated a split vaccine. Since such vaccines are safe and efficient, they have been in use for many years to protect man against enveloped RNA viruses, e.g., influenza and measles. Seven rhesus monkeys were immunized at Week 0, 4, 8, and 16 by im injection of 2 ml of vaccine containing 140 micrograms of Tween-ether-disrupted SIVmac251/32H adsorbed onto aluminum hydroxide. The immunized animals and three nonvaccinated control monkeys were challenged 2 weeks after the last immunization by iv injection of 10 to 50 minimal monkey infectious doses of SIVmac251/32H. Four of seven immunized animals did not show any signs of virus replication and therefore appeared to be protected. Nonvaccinated control animals and the vaccine failures showed a rise in their urinary neopterin concentrations 1 to 2 weeks after infection. At the end of the second week and thereafter, cocultures and polymerase chain reaction of their peripheral blood lymphocytes were positive. After the challenge, control animals and infected vaccinees showed a primary or secondary antibody response while antibody titers declined in virus-negative animals. Specific cytotoxic T-lymphocytes were not present prior to challenge, but were present in some animals thereafter. Therefore, these seem to reflect a response to viral replication rather than to immunization. Prior to challenge the CD4-positive lymphocytes of the peripheral blood of the four virus-negative animals only proliferated after exposure to the immunizing antigen. Thus, this reaction appears to predict protection.