We have previously studied the relationship between social subordinance (by approach-avoidance criteria) and physiology among male olive baboons (Papio anubis) living freely in a national park in Africa. In stable hierarchies, subordinate individuals have elevated basal glucocorticoid concentrations and a blunted glucocorticoid response to stress, as well as a prompt suppression of testosterone concentrations during stress. These facets have been interpreted as reflecting the chronic stress of social subordinance. In the present report, we find these endocrine features do not mark all subordinate individuals. Instead, endocrine profiles differed among subordinate males as a function of particular stylistic traits of social behavior. A subset of subordinate males was identified who had significantly high rates of consortships, a behavior usually shown only by high-ranking males. Such behavior predicted the beginning transition to dominance, as these males were significantly more likely than other subordinates to have moved to the dominant half of the hierarchy over the subsequent 3 years. In keeping with this theme of emerging from subordinance, these individuals had also significantly larger glucocorticoid stress-responses, another feature typical of dominant males. However, these subordinate males also had significantly elevated basal glucocorticoid concentrations; it is suggested that this reflects the stressfulness of their overt and precocious strategy of reproductive competition. In support of this, subordinate males with high rates of covert "stolen copulations" did not show elevated basal glucocorticoid concentrations. A second subset of subordinate males were the most likely to initiate fights are to displace aggression onto a third party after losing a fight. these males had significantly or near-significantly elevated testosterone concentrations, compared to the remaining subordinate cohort. Moreover, these males had significantly lower basal glucocorticoid concentrations; this echoes an extensive literature showing that the availability of a displacement behavior (whether aggressive or otherwise) after a stressor decreases glucocorticoid secretion. In support of this interpretation suggesting that it was the initiation of these aggressive acts which attenuated glucocorticoid secretion, there was no association between glucocorticoid concentrations and participation (independent of initiation) in aggressive interactions. Thus, these findings suggest that variables other than rank alone may be associated with distinctive endocrine profiles, and that even in the face of a social stressor (such as subordinance), particular behavioral styles may attenuate the endocrine indices of stress.