Demand estimation for a set of goods is usually an integral part of the planning process. The forecasts should bring out the likely gaps in demand and supply which, in turn, should facilitate the planned enhancement of supplies/or formulation of policies for demand management. This paper review various demand forecasts of household energy up to 2005, and brings out that there has been substantial variations, even inconsistencies, in the estimates of future demand by various agencies. The inconsistencies have been more prominent for fuelwood demand forecasts: from less than 100 to more than 300 million tonnes by the turn of the century. Such differences have been a result of variations in the number and nature of variables, assumptions, analysts' perceptions with respect to marginal costs of different fuels, and the interpretation of the term demand itself. It is concluded that considerable time series and cross sectional data must be collected, processed, updated and shared so that the variations in energy demand forecasts can be minimized. In the meantime, realistic estimates of availability and fuelwood production potential of forest and non-forest wasteland must also be attempted to arrive at a practicable and desirable mix of different fuels to meet the total household energy demand.