We argue that consumer credit spreads matter materially for household choices and that time-varying spreads have important distributional consequences. Studying Danish household data, we show that elevated consumer credit spreads reduce indebted households’ consumption and that the marginal propensity to consume is countercyclical partially due to credit spreads. We study a HANK-model in which banks provide consumer credit and corporate loans. Through countercyclical credit spreads, frictional finance amplifies aggregate shocks and induces consumption inequality. Economies with less leveraged banks may experience reduced aggregate volatility by muting the financial accelerator, but may also face higher volatility and lower welfare at the household level.