This dissertation focuses on two themes in the field of empirical banking. First, the importance of credit relationships for corporations experiencing financial distress; second, the relevance of credit market structure for the price and non-price terms associated with the provision of credit. The first chapter investigates whether relationship lending helps firms experiencing idiosyncratic financial distress. By constructing a novel dataset on syndicated lending that tracks the availability and pricing of credit for US corporate borrowers over three decades, I conclude that relationship lending benefits borrowers in distress. Specifically, I find that relationship lenders provide a higher credit amount, charge lower interest rates, and require similar collateral and fees, relative to non-relationship lenders. Firms benefit from relationship lending irrespective of their access to outside financing options. Overall, my findings provide support to theories of implicit commitment and reputational capital in lending relationships. The second chapter focuses on the importance of bank specialization in monitoring specific projects for the design of loan contracts. Using detailed information on credit relationships in the US syndicated loan market, I first document that banks specialize in lending to specific industries. Specialization is common across industries and persistent over time. Then, I provide evidence that specialized banks provide credit with looser restrictions (covenants) and at lower cost (interest rates) when arranging loans in their sector of specialization. This is consistent with an explanation of bank specialization based on monitoring advantages. Overall, the laxer contract terms offered by specialized banks could provide an explanation for recent evidence that firms cannot easily substitute credit granted by specialized banks.