Using a linked database of Paycheck Protection Program (PPP) loans and Yelp-listed restaurants, we document that businesses owned by minority racial groups are more likely to use fintech lenders than traditional lenders. We develop a simple two-sided matching model to show that this phenomenon can be potentially attributed to differences in performance among borrowers, racial disparities in lending relationships, and race-dependent values of borrower-lender matches. Empirically, we do not find consistent evidence that operational performance is an explanation. We find that minority-owned restaurants are less likely to have lending relationships and that restaurants without lending relationships are more likely to use fintech lenders. We also find a more negative minority-non-minority gap in operational performance for fintech lenders, suggesting minority-owned businesses have higher matching values with fintech lenders. We do not find a similar pattern for first-time bank participants, community development financial institutions, credit unions, or other non-federally insured lenders. Overall, our results suggest that there are racial barriers in traditional loan distribution channels and this can be at least partially addressed by fintech lenders.