Earlier attempts on forecasting energy demand for agriculture have largely been based on aggregate macro-level data and overall national averages and trends. Useful as they are in indicating some broad trends, these forecasts are unable to capture variations between crops, regions, operations and levels of technology used. The disaggregated energy demand forecasts are thought to be more meaningful for policy decisions. The present study is limited to only commercial or conventional form of energy, that is electricity and petroleum products. Also, the energy demand estimations are restricted to six major crops, viz., paddy, wheat, sugarcane, cotton, oilseeds and pulses. In terms of energy inputs into agriculture, the demand estimations are made in relation to land preparation, fertilizer (NPK), pumping irrigation, harvesting (including transport from farm to village) and threshing (relevant for paddy and wheat only). The percentage increase in energy consumption required to achieve 1% increase in crop production by the year 2000 turns out to be 2.1% for paddy, 1.4% for wheat, 2.2% for oilseeds, 7.9% for pulses, 1.6% for cotton and 5.5% for sugarcane. Thus, the output of wheat and cotton is likely to show much higher response to increase in energy inputs as compared to other crops. Moreover, significant pressure in terms of increased energy requirements would be created by an attempt to accelerate the growth of production of sugarcane and oilseeds through a policy designed for achieving major transition from traditional agriculture to modern agriculture.