In this paper we test how much party labels influence vote choices between candidates with realistic distributions of candidate positions. We use the UK Parliamentary Candidates Survey to generate realistic distributions of political positions and demographic attributes for Labour and Conservative candidates. We then create electoral matchups between randomly selected Conservative versus Labour candidates, with half of respondents seeing party labels in addition to candidate positions and demographics and half not seeing party labels. We analyse the relationship between respondents’ own positions and characteristics and those of candidates in the no party labels condition to understand which dimensions of candidate-respondent similarity are most highly weighted in decisions. We also compare these relationships as well as aggregate vote choice in the no party labels versus party labels conditions to understand the effect of party cues, and how they modify candidate selection in a context where candidates have realistic patterns of difference. We find evidence that party cues have much larger effects on the vote choices of past Conservative voters than on those of past Labour voters. The past Conservative voters are substantially more likely to choose the Conservative candidate in the experiment when the party label is present, whereas there is negligible effect on past Labour voters’ support for the Labour candidate in the experiment when the party label is present versus when it is not