Laboratory experiments involving a real effort task are conducted to examine the importance of gender differences in competition aversion for generating gender wage gaps. Cross-subject design treatment and control experiments suggest that gender differences in risk aversion play no significant role in competitive (tournament) vs. piece-rate job choices and consequent gender wage gaps. Subjects in the treatment experiments are sorted into relatively more and relatively less risk averse groupings. Relatively less risk averse subjects are assigned to a risky job track involving a known constant probability of unemployment in each period. The gender wage gap contribution of gender differences in competition aversion compared with the contribution of gender differences in performance is especially large for relatively less risk averse subjects.