Using a unique data set of classified ads in South Africa, I explore whether employers discriminate against immigrants in the hiring process. I develop a quasi-experimental method to estimate discrimination exploiting variation in the applicant pool composition due to the timing of postings. Consistent with a tournament models in which immigrants are penalized, I find that both foreigners and natives benefit from being pooled with foreign job seekers. Next, I test whether discrimination affects search behavior. Controlling for location fixed effects, I find suggestive evidence for sorting: immigrants search further away and higher discrimination in the residential area is positively correlated with the decision to search in different suburbs. This additional cost to job seekers has not been explored in the discrimination literature.