Research on testosterone-behavior relationships in humans is assessed in relation to a version of the challenge hypothesis, originally proposed to account for testosterone-aggression associations in monogamous birds. Predictions were that that testosterone would rise at puberty to moderate levels, which supported reproductive physiology and behavior. Sexual arousal and challenges involving young males would raise testosterone levels further. In turn, this would facilitate direct competitive behavior, including aggression. When males are required to care for offspring, testosterone levels will decrease. Testosterone levels will also be associated with different behavioral profiles among men, associated with life history strategies involving emphasis on either mating or parental effort. Most of these predictions were supported by the review of current research, although most studies were not designed to specifically test the challenge hypothesis.