Agricultural support payments that cause no or minimal production distortions are exempt from World Trade Organization restrictions. If and how much decoupled payments, such as direct payments in the U.S., affect agricultural production remains an open empirical question with implications for policy. We use multiple years of the Census of Agriculture to estimate the aggregate supply response to changes in direct payments. To identify an exogenous source of variation in payments we exploit a provision of the 2002 Farm Act that departed from previous policy by making oilseeds eligible for direct payments, thus increasing payments to areas that historically produced more oilseeds. Using a sample of ZIP codes that accounts for more than eighty percent of the national value of production of program crops, our instrumental variable estimates, in contrast to OLS estimates, suggest that changes in payments over the period 2002 to 2007 had little effect on aggregate production.