To explain women's underrepresentation in politics, supply-side factors receive much empirical support, emphasizing the low numbers of women on the ballot. Whether demand from voters also contributes to the problem is less clear, however, as both observational and experimental research shows that average voters are not less likely to vote for women candidates. We argue that voters actually do play a role, although not all voters to an equal extent. More precisely, we expect the gender bias in the electorate to be conditional upon partisanship and propose two mechanisms through which this materializes: political gender attitudes and/or gender stereotypes. Although the conditionality of voters' gender bias based upon partisanship is convincingly shown to exist in the US, much less is known about it in the European context, while its multi-party political systems lend themselves well for a more detailed differentiation between party families. We expect that right, and especially populist radical right, voters are biased in favor of men politicians, while left, and especially green left, voters are biased in favor of women politicians. We test our hypotheses with a large-scale vignette experiment (N = 13,489) in the Netherlands, and show that there is indeed a (slight) preference for women representatives among Green party voters, and a clear preference for men candidates among voters of populist radical right parties. Moderate left-wing or right-wing voters, however, show no gender bias. Thus, although right-wing populist parties have electoral incentives to be hesitant about promoting women politicians, most other parties face no electoral risk in putting forth women politicians.