This dissertation examines whether non-co-ethnic candidates can benefit from shared ethnicity. Using the test case of non-Hispanic White candidate appeals to Latino voters, I argue that non-Latino candidates who use a package of appeals I call identity-based appeals can reduce the social space between themselves and Latinos and increase support. The first chapter consists of four candidate case studies based on news accounts, campaign materials, and archival records. These case studies illustrate the varied uses of identity-based appeals across states, political parties, and time. The second chapter uses observational vote data to conduct ecological inferences analyses for candidates from the first chapter. Evidence suggests that relative to comparison cases who do not use identity-based appeals, candidates who use identity-based appeals perform better among Latino voters. The final chapter uses a novel conjoint experiment to examine the relative impact of each identity-based appeal as well as compare the effectiveness of identity-based appeals to policy and partisan cues. While evidence suggests identity-based appeals are effective at increasing candidate support among Latinos, policy appeals on immigration have significantly larger effects. Evidence suggest political knowledge moderates the effects of identity-based appeals.