The sexual relationships of 15 adult male rhesus macaques (Macaca mulatta), of one social group in the Cayo Santiago colony, Puerto Rico, were studied during the 1981 mating season. Two criteria were used to determine whether or not a focal male was in consort in a given 20-min observational sample. One hundred and thirty-two consortships were recorded. The distribution and duration of all consortships, and the distribution of those consortships that coincided with the estimated time of conception, were positively correlated with male dominance rank and length of tenure. Correlations with dominance were stronger than those with tenure. Older females had more consortships with focal males than younger females. There was no relationship between female rank and the distribution of consortships. Consortships did not give exclusive access to receptive females although they may do in feral situations. Males were primarily responsible for maintaining proximity in about two thirds of consortships. The four top-ranking males were primarily responsible in all but one of their consortships. Almost all of those in which the female was primarily responsible involved younger, slightly lower-ranking males. This pattern may have resulted from the females being attracted to the latter males although other interpretations are possible. The partner who was primarily responsible for maintaining proximity also tended to be the predominant groomer, supporting the view that grooming plays a role in the maintenance of sexual consortships. Almost a third of consortships were with females who had conceived, suggesting that males could not accurately assess female reproductive state. Consortships maintained by the male were longer than those maintained by the female. Males may try to increase their chances of fertilising a female by prolonging the consortship. High-ranking males may have been more successful at this, resulting in the positive correlation between male dominance rank and consortship duration.