The Challenge Hypothesis postulates that androgen levels are a function of the social environment in which the individual is living. Thus, it is predicted that in polygynous males that engage in social interactions, androgen levels should be higher than in monogamous animals that engage in parental care. In this study, we tested this hypothesis at the intra-specific level using a teleost species, Sarotherodon galilaeus, which exhibits a wide variation in its mating system. Experimental groups of individually marked fish were formed in large ponds with different operational sex-ratios (OSR) to study the effects of partner availability on blood plasma levels of sex steroids [11-ketotestosterone (11-KT), testosterone (T), and 17,20beta-dihydroxy-4-pregnen-3-one (17,20beta-P)] and gonadosomatic index (GSI). Polygyny mostly occurred in the female biased OSR groups. 17,20beta-P and gonadosomatic index did not differ among OSR groups. However, 11-KT was high in male biased OSR and positively correlated with aggressive challenges, thereby supporting the central postulate of the Challenge Hypothesis. The results of T were the inverse of those of 11-KT, probably because 11-KT is metabolized from T. 11-KT levels of polygynous males did not differ neither from those of monogamous males, nor from those of males that participated in parental care. These results do not support the expected relationships between polygyny, parental care, and androgen levels. The differences from expectations for 11-KT may be related to the fact that in S. galilaeus, the mating and the parenting phase are not clearly separated and thus, males may still fight and court while they are brooding.