Plantation managers use a variety of decision-support systems to assist in deciding oil optimum silvicultural treatments. Often these include computer-based growth simulation models that include an economic analysis. We recommend that energy analyses should be added as an additional tool, to ensure that fossil and other energy is used wisely and to meet other objectives such as reducing environmental impacts. In this study, selected silvicultural inputs were evaluated for hypothetical Eucalyptus grandis and Pinus taeda plantations. Silvicultural treatments resulted in very large differences in energy output: energy input ratios, with very high ratios being associated with treatments that result from a change of site quality by overcoming major soil limiting factors. Machinery choice and other material inputs such as fertilizers and herbicides also influenced energy balance ratios. The use of nitrogen fertilizers was an energy-intensive option. The coppice eucalyptus plantation often had higher energy balance ratios than the pine because of its faster growth rate and higher basic wood density. Evaluating a combination of silvicultural treatments was not as helpful for optimizing energy balance ratios as evaluating them separately. Yield responses vary widely with species, sites and a range of operational factors. Thus determining the yield response required for a silvicultural treatment to meet a prescribed energy output: input ratio would allow managers to quickly evaluate treatments. In this Study, We used an energy balance ratio of 25 as a criterion. Some treatments would have met this standard while others were marginal or the responses could not meet the criterion. However, the selection of the critical ratio needs further study. Net energy yield per hectare, evaluated for various silvicultural options, would be a useful measure in planning energy-use systems for forest estates that supply bioenergy industries. It was a less sensitive measure than energy balance for comparing different silvicultural treatments. (c) 2005 Elsevier Ltd. All rights reserved.