The challenge hypothesis proposes that testosterone (T) elevation above what is needed for breeding is associated with social factors, and males possibly modulate their hormonal response to variations in population density and sex ratio. We investigated the role of social environment in altering testosterone levels and aggression in a tropical, seasonally breeding grassquit (Volatinia jacarina). We exposed males to three social conditions during 1 year: all-males treatment (six males), mixed treatment (three males-three females), and paired treatment (one male-one female). We quantified aggressiveness among males and T plasma concentration for each individual in each treatment monthly. We found that more aggressive interactions occurred in the all-males treatment than in the mixed treatment. The data also revealed that, coincident with these behavioral changes, the patterns of T variation through time in each treatment were markedly different. The all-males treatment exhibited an early increase in T concentration, which was sustained for a lengthy period with two distinctive peaks, and subsequently declined sharply. The mixed treatment presented an intermediate pattern, with more gradual increase and decrease in T levels. At the other extreme, the paired treatment presented a later rise in T concentration. We conclude that the more competitive environment, with higher density of males, caused the early and higher elevation in T level, thus the presence of competitors may influence the decision of how much a male should invest in reproduction. We suggest that the male's perception of his social environment ultimately mediates hormonal production and alters his reproductive strategy.