Summary. Six fertile and healthy adult Macaca fascicularis males were studied. Radioimmunological assay of the plasma testosterone, sampled without anesthesia in the afternoon at the beginning of each month, showed an annual hormone cycle with a maximum (16.7 ! 1.1 ng/ml) in the fall and a minimum (9.5 ! 0.9 ng/ml) in the spring ( fig. 1). Using the ratio : number of ejaculations/number of trials, the ejaculatory ability of these animals ( fig. 2) was estimated for 10 min in the presence of females between days 12 and 15 of their menstrual cycle. This ability showed no cyclic variation during the year.Comparing the annual variation of testosterone levels in macaque males (Macaca mulatto, Macaca arctoides, Macaca nemestrina, Macaca fascicularis) and man, we found that, except for Macaca arctoides and Macaca nemestrina, the maximal simian levels always coincided with autumn and the minimal levels with spring in spite of the different rearing environments. (table 2).After studying ejaculatory ability and plasma testosterone level in the intact macaque throughout the year and comparing it to the results obtained by Resko and Phoenix (1972), Phoenix et al. (1973) and bVilson (1974, 1975) studying castrated males, we believe that above a minimal level, variation in the plasma testosterone level does not affect male sexual behavior, at least as far as ejaculatory ability is concerned. Moreover, du ring this study we noted that above a maximal plasma testosterone level, varying with the season, the ejaculation rate may be depressed ( fig. 3).Introduction.