Two problems make estimation of import demand for developing countries extremely difficult. The first is a conceptual problem: how to model government intervention which simultaneously distorts consumer and producer prices and foreign exchange allocations. The second problem is a technical one: how to obtain econometrically consistent estimates of an import demand equation which includes all the major parameters, given only 10 years or less of time series data. Both of these problems are addressed in this paper which presents an attempt to estimate Mexican corn import demand. Results indicate that Mexican corn imports could be substantially reduced by transferring relatively small amounts of subsidies from consumers to producers. Furthermore, effects on Mexican corn import demands of world border prices, exchange rates, and foreign exchange allotments are insignificant.