This paper presents food demand elasticity estimates for thePakistan by urban and rural populations by income quartile for thirteenfood groups. Income and price elasticities are estimated using a newfood demand estimation technique based on demand for characteristics.This new technique requires far less data than the usual econometricapproaches and so may be implemented relatively quickly andcosteffectively. However, the resulting demand elasticity estimatesdepend directly on strong a priori assumptions made concerning fooddemand behaviour, but assumptions which do not depend on assumptions ofweak or strong separability. Rather quite the opposite assumption ismade – that the marginal rate of substitution’> between two foodsdepends directly on the levels of consumption of all other foods. Thepaper is organised as follows. The first section provides a briefoverview of the methodology used for undertaking the demand estimations.The second section discusses the-data used for these estimations andfood consumption patterns. The third section presents the demandelasticity estimates.