Variation in human male testosterone levels may reflect, and effect, differential behavioral allocation to mating and parenting effort. This proposition leads to the hypothesis that, among North American men, those involved in committed, romantic relationships will have lower testosterone levels than men not involved in such relationships. Our study is the first to examine whether being in such a relationship (rather than being married) is the meaningful predictor of male testosterone levels. To test this hypothesis, 122 male Harvard Business School students filled out a questionnaire and collected one saliva sample (from which testosterone level was measured). Results revealed that men in committed, romantic relationships had 21% lower testosterone levels than men not involved in such relationships. Furthermore, the testosterone levels of married men and unmarried men who were involved in committed, romantic relationships did not differ, suggesting that, at least for this sample, male pair bonding status is the more significant predictor of testosterone levels than is marital status.