Many hypotheses have been proposed to explain why women remain underrepresented in Congress. One of those hypotheses is that some voters have blatant prejudices against women politicians, while others hold stereotypes about men and women politicians that favor men. In contrast, others claim that women candidates for Congress actually have an advantage in running for office because voters prefer women politicians. We test those hypotheses using pooled 1988, 1990, and 1992 National Election Studies data and the pooled 1988-1992 Senate Election Study and building on Krasno's (1994) model of voter choice in House and Senate elections. We find evidence that some voters prefer women candidates in House races, but not in Senate races. The advantage for women candidates in races with a challenger and incumbent is slight and can be attributed to the strong support of well-educated women voters. An advantage for women candidates is more pronounced in open-seat contests. In open-seat races, women voters, regardless of their education levels, more strongly support women candidates. Overall, candidate sex was not significant to male voters.