Many low-income countries pursue cheap-food policies in which consumers pay subsidized prices for bread, rice and other staples. This paper addresses the issue of why different governments select different food subsidy policies, using multiple instruments rather than a simple across-the-board subsidy to provide consumers with access to cheap food. It examines the optimal structure of cheap-food policies in the context of a partial equilibrium model in which the country may be large in trade, and is able to combine import subsidies or tariffs, and output taxes or subsidies, to transfer income to consumers through the market. The model allows for a marginal opportunity cost of government revenues greater than one dollar. In addition, in the model, food aid from overseas may be either given away to the consumer, or given to the government for subsequent sale in the domestic market. The results indicate that only by happenstance will a country choose to use a pure consumption subsidy or a pure import subsidy to transfer income to consumers. In addition, an increase in international food aid does not necessarily lead the government to reduce producer and consumer prices for a commodity.© 1999 Elsevier Science B.V. All rights reserved.