We document that the quasi-mandatory U.S. flood insurance program reduces mortgage lending along both the extensive and intensive margins. We measure flood insurance mandates using FEMA flood maps, focusing on the discreet updates to these maps that can be made exogenous to true underlying flood risk. Reductions in lending are most pronounced for low-income and low-FICO borrowers, implying that the effects are at least partially driven by the added financial burden of insurance. Our results are also stronger among non-local or more-distant banks, who have a diminished ability to monitor local borrower adherence to complicated insurance mandates. Overall, our findings speak to the unintended consequences of (well-intentioned) regulation. They also speak to the importance of factoring in affordability and enforcement feasibility when introducing mandatory standards.