Rapid growth in the demand for commercial energy in Kerala, a state in India, posed serious development constraints in the recent past, The crucial issue of managing this demand will be of great importance in formulating the future development policy of the state. This study analyses the requirement of three major forms of commercial energy in the state of Kerala (viz. electricity, petroleum products and coal). Sectorwise/productwise econometric demand models are generated using regression method. The primary regression models consisting of statistically significant key energy indicators have been fitted with ordinary least-squares (OLS) estimation. Models wherever required have been subsequently refined in the second stage computations by employing Cochrane-Orcutt transformation algorithm to remove the effects of auto-correlation. The accuracy of the models have been checked using standard statistical techniques and validated against the past data by testing for 'expost' forecast accuracy. Employing the models, medium-range projections of demand of the state for commercial energy till the year 2020 have been made. The policy implications and related key issues have been addressed. The study identifies the urgent need for special attention in evolving effective energy policies to alleviate an 'energy famine' in the near future.