Is monetary policy less effective at increasing real output during periods of high volatility than during normal times? In this article, I argue that greater volatility leads to an increase in aggregate price flexibility so that nominal stimulus mostly generates inflation rather than output growth. To do this, I construct price-setting models with "volatility shocks" and show these models match new facts in CPI micro data that standard price-setting models miss. I then show that these models imply that output responds less to nominal stimulus during times of high volatility. Furthermore, because volatility is countercyclical, this implies that nominal stimulus has smaller real effects during downturns. For example, the estimated output response to additional nominal stimulus in September 1995, a time of low volatility, is 55% larger than the response in October 2001, a time of high volatility. JEL Codes: E10, E30, E31, E50, D8.