We develop and estimate a general equilibrium signaling game model of xenophobia, in which the reputational consequences of committing xenophobic acts are determined endogenously. Using our unique survey data, which enables the identification of the model, we find that reputational incentives are a key driver of xenophobic behavior, suggesting that policies targeting these motivations could be effective. We then apply this structural model to assess the impact of the COVID-19 pandemic on anti-Chinese xenophobia. Our findings suggest that the majority of the post-pandemic rise in xenophobic behaviors is driven by changes in reputational incentives.