Using data from U.S. states, this paper examines empirically how the effects of monetary policy on output depend on banking conditions. I find that when a state’s banking sector starts out with a low capital-asset ratio, its subsequent lending and output growth are both more sensitive to changes in the federal funds rate or other indicators of monetary policy. This result is consistent with the existence of a ‘bank capital channel’ as well as a conventional bank lending channel. Other evidence favors the capital channel, as bank liquidity is not associated with variation in the impact of monetary policy on output at the state level. Consistent with the theory, the importance of state-level measures of bank capital weakens significantly after the removal of regulatory barriers to interstate banking.