Farmers’ decisions about how much crop insurance to buy are not generally consistent with either expected profit or utility maximization. They do not pick coverage levels that maximize expected subsidy nor do they demand full insurance coverage. In addition, the absolute size of farmer-­‐paid premium seems to influence the type of insurance product farmers buy. Understanding demand drivers for crop insurance has taken on new importance because of the expanded role Congress has designated for crop insurance as a key part of Federal farm policy. By modeling financial outcomes as gains and losses, prospect theory offers an appropriate framework to better understand farmers’ purchase decisions. Because insured events are best modeled as continuous random variables, cumulative prospect theory is used to find a theoretical foundation that can explain farmers’ anomalous decisions. The role of the reference point that defines outcomes as either a gain or a loss, the degree of loss aversion, and the probability weighting function are explored under typical distributions of price, yield, and revenue for a corn producer. Choice of reference points that are consistent with farmers using crop insurance to manage risk are not consistent with observed purchase decisions. Choosing the reference point to make crop insurance akin to a stand alone investment generates optimal choices that are consistent with observed decisions and with the way that insurance agents sell the product.