Olfactory influences are important in the control of reproductive function in many mammalian species including primates. Previous studies have shown that a mixture of volatile aliphatic acids (copulins), normally present in the vaginal secretions of rhesus monkeys, can exert a stimulatory effect on the sexual behaviour of males when pairs are tested in small cages. The present study deals with the role of these substances when tested in social groups of rhesus monkeys housed in a large testing arena, the purpose being to assess the effectiveness of olfactory signals under more nearly naturalistic conditions. Each group consisted of one male with four long-term ovariectomized females from which the male could choose. A counter-balanced experimental design was used in which each female of a testosterone-treated pair was given applications to the sexual skin of either copulins or control ether immediately before each 1-h behaviour test. After 16 tests, the procedure was repeated with the previously untreated pair in each group, and nine groups were separately studied. Testosterone was used here simply to enhance female proceptivity. The results showed that females received significantly more ejaculations (P less than 0.05), mounting attempts (P less than 0.01) and mounts (P less than 0.05) when bearing copulins than when bearing control ether. There were positive responses to copulins in 12 of the 36 males-females pairs, involving all males, and positive responses to ether in five of 36 pairs, involving four males. Copulins resulted in a change in the choice of sexual partner in five males, but there were no changes in the choice of partner with ether. Social factors, such as dominance, could override the responses to olfactory signals, and there was a significant treatment order effect, but this was eliminated by the counter-balanced design. These findings indicate a significant role for olfactory communication in mate selection in a socially living higher primate.