This paper examines the role of energy in long-term growth. The authors estimate a translog production function in capital, labour, and energy, using data on 38 countries from 1965-1990. It is found that separating the countries into two groups, the high growth group and the rest of the world (ROW), yields slope coefficients that are statistically different. Furthermore, there are dramatic differences in the estimated production elasticities, returns to scale, and bias in technical change for the two country groups. Based on the estimated coefficients, the authors approximate the contributions ofthe factors ofproduction in explaining the observed growth in real GDP through a growth accounting exercise. Energy emerges as an important source of growth for both country groups, second only to increases in total factor productivity. In light ofthe negotiations over a Protocol to the Climate Change Convention, the authors estimate the capital and labour requirements for national energy reductions. Finally, the authors use the production function to provide analytical expressions for the elasticity of energy intensity with respect to other factor inputs, and also for autonomous energy efficiency improvements (AEEI). It is found that the deceleration ofthe AEEI in the ROW countries more than offsets the combined acceleration witnessed in Hong Kong, South Korea, and Thailand. -