The credit channel of monetary policy states that informational frictions in credit markets worsen during tight monetary policy or economic downturns. These frictions show how bank balance sheet weakness affects banks’ ability to lend (Bank Lending Channel), whereas firm balance sheet weakness affects firms’ ability to borrow (Balance Sheet Channel). In the cross-section, the bank lending channel predicts that constrained banks reduce lending (increase rates) in aggregate more than non-constrained banks. Whereas the balance sheet channel predicts that banks reduce lending (increase rates) to weaker firms in aggregate more than stronger firms. In this paper, we empirically investigate the relative importance of each of these frictions with respect to the interest rates paid by firms. Whereas most studies analyzing interest rates use bank level averages, we have information at the bank-firm-loan level. Studies using bank level interest rates, however, do not allow the identification of each of the above frictions. For example, in the case of assortative matching between banks and firms (i.e. weak banks lend relatively more to weak firms) changes in interest rates at the bank level can be both the product of bank balance sheet effects and firm balance sheet channel effects. Therefore, what appears as a bank balance sheet effect may in fact be due to a firm balance sheet effect. In our study, by using loan level data matched with bank and borrower characteristics, we can identify both bank and firm balance sheet channel effects.