Testosterone levels of 59 male rhesus monkeys were monitored over a period of 5 years. Longitudinal comparisons revealed consistent rises in mid-morning levels of circulating hormone in successive years from age 2.5 to 6.5 years of age, whereas cross-sectional comparisons failed to detect significant differences among the older subjects. The first mid-morning hormonal elevation could be detected in some males as young as 2.5 years of age, whereas other males showed no detectable rises until age 5.5 years. Males showing first rises at later ages did not show hormonal levels consistently below age peers who had shown earlier rises. Extreme month-to-month variability and a failure to manifest the seasonal normal curve of fully adult males was characteristic of younger males, but some of these males, nonetheless, proved capable of fertilizing females. Although hormonal and agonistic dominance measures failed to show consistent correlations, the alpha male in an age cohort significantly more often had the highest testosterone levels. These data are used to argue that adolescence is a process that takes place over several years and that classification of adolescent animals as adults, based on a single criterion like fertility, has confounded many prior studies involving cross taxa comparisons as well as developmental variables.