We analyze the effect of the ECB's Quantitative Easing program (Expanded Asset Purchase Program - EAPP) on bank lending using security-level bank balance sheet data combined with a comprehensive dataset on new loans in Portugal. Our identification relies on the fact that only a subset of Portuguese banks was exposed to EAPP via prior holdings of EAPP-eligible securities and origination of eligible ABS and covered bonds. Using a difference-in-differences specification with borrower and bank fixed effects, we find that lending rates to the same borrower drop by 64 b.p. at banks exposed to QE relative to banks not exposed to QE. Loan volumes to existing corporate clients grow by one percentage point faster at exposed banks relative non-exposed banks. This result is robust to including both bank and borrower*time fixed effects, as well as a wide range of loan and borrower characteristics. At the extensive margin, the probability of credit approval to a new corporate client is about 1 percentage point higher at exposed banks post-QE announcement.