P C Bansil IN their recent paper 'Foodgrains Demand in India to 2020: A Preliminary Exercise', G S Bhalla and Peter Hazell (Economic and Political Weekly, Review of Agriculture, December 27, 1997) have projected feed and food demand for India till 2020 under different scenarios, They have added a few new dimensions to the discussion and analysed the data on the basis of comprehensive studies done at IFPRI. The authors have accepted that "in the past, cereal demand has not grown as rapidly in recent years as many experts had previously expected. In urban areas, per capita cereal consumption seems to have stabilised at about 135 kg per year, while it has actually declined in rural areas (from 185 kg to 175 kg per year) during the last two decades". We do not know the source of these numbers which according to NSSO estimates arc not only higher than the consumption pattern during 1993-94, but also during 1987 88 (Table I). The study is based on the elasticities for human consumption given in their Tabic 2, feed coefficients in their Table 4 and baseline (1990) availability estimates in column I of their Table 5. All these three assumptions call for comments.