This study estimates a complete demand system for rural Bangladesh, applying the Almost Ideal Demand System (AIDS) model. The estimates of demand parameters are based on primary data from the rural household survey conducted by IFPRI in 1991/92. Even though the sample may over-represent the rural poor, the estimates of income elasticity of demand suggest that rural households in general are highly responsive to changes in income in adjusting their consumption patterns. Demands for commodities are also quite responsive to changes in their own prices, with the exception of salt. The estimates of cross-price elasticities indicate that substitution effects are strong, and therefore have important implications for price policies. Disaggregated by income groups, the estimates of demand parameters of rice and wheat suggest that low-income households are more price ad income responsive than high-income households are more price and income responsive than high-income households. Differences in elasticities in absolute values between the two groups are quite striking. The estimates conform with the findings of other studies that wheat is an inferior commodity in rural Bangladesh. This attribute makes wheat a self-targeting commodity for targeted food intervention programmes. The study contends that government price intervention programmes. The study contends that government price interventions may lead to serious price repercussions in the economy. In contrast, income generating programmes and policies foster higher levels of consumption for all normal Commodities, and thus, a steady growth in production by enhancing effective demand.