To what extent and under what conditions does candidate gender impact voter choice(s)? While research has focused on measuring individual and aggregate voter biases, it is less clear how these are distributed within the electorate. In this research note, we argue that the degree to which voters’ preferences are biased by candidate gender may vary across voters and votes. We show that the increasing availability of individual-level ballot data opens up new opportunities to study the distribution of group affinity voting within electorates. Drawing on simulations and real data from Lithuanian national elections, we demonstrate that [mixed logit models/social network analysis] can be used to investigate the distribution of gender biases across and within individual voters and votes, and to assess the relative importance of gender compared to other candidate characteristics. We thereby offer a new approach for the comparative study of affinity voting that can be applied to a variety of electoral systems and identity groups.