Political polarization in the United States has risen dramatically in recent years. A growing body of work shows that partisan divides have increasingly bled into nonpolitical domains, including friendships and romantic relationships. Recent surveys suggest that prejudice based on party identity now exceeds bias directed at racial and religious out-groups. This study uses a conjoint survey experiment to assess the importance of shared partisanship for long-term partner preferences. We asked a national sample of unmarried young adults (N = 1483) to evaluate profiles of fictional romantic partners who varied in regard to party identification, religion, race/ethnicity, education, and other factors. Our experiment allows us to address how party identity is distinct from other potential partner traits routinely examined in studies of assortative mating (race/ethnicity, religion, and education). The results of our experiment suggest that individuals prefer potential romantic partners who share their party identity. Among partisans, this preference is primarily attributable to negative partisanship, that is, an aversion to those in the opposing party. Focusing on white partisans, we find that Democratic women are more likely than any other gender/party group to treat out-party membership as a relationship deal-breaker. We discuss the implications of these findings for knowledge on mate preferences and future research on this topic.