Many governments in developing countries distribute fertilizer at subsidized prices to stimulate fertilizer use and agricultural productivity. However, the extent to which government fertilizer subsidy programs affect private suppliers' sales of fertilizer is largely unknown and subject to considerable debate. This paper models farmer input use decisions in a two-channel marketing system and develops an econometric method to measure the potential "crowding out" effect of government subsidy programs. Using nationally representative household panel survey data in Zambia, we estimate how household fertilizer acquisition from commercial dealers and government programs are influenced by household socioeconomic, demographic, and market variables, and agro-ecological factors. Results indicate that the crowding out effect is sensitive to where governments choose to operate subsidy programs. The quantity of fertilizer procured from private sector may fall substantially if government chooses to distribute fertilizer in areas where fertilizer use is likely to be profitable and where private retail networks are already operating. On the other hand, the crowding out effect is negligible if government subsidy programs operate in areas where private fertilizer sales are low; in such areas, government programs may possibly enhance commercial demand over the long run.