Energy forecasting is an important task of immense value for predicting the future development and implementation of energy technologies. While a certain degree of uncertainty is to be expected in foreseeing future conditions, energy forecasts are typically quite inaccurate in their predictions. The possible causes of such inaccuracy are examined in this paper in the context of an energy forecast study conducted more than 30 years ago. This study, conducted using the Delphi technique, provided predictions for the development of nuclear energy production, nuclear reactor safety, fossil energy, energy transmission, environmental effects and the renewable energy in the years 1985 and 2000. A comparison with the current progress of the energy industry indicates that the predictions of the study were highly optimistic, and most of the predictions did not materialize nor are likely to be realized in the near future. An analysis of the probable causes of failure is presented in this paper. Such analysis can provide a better understanding of the forecasting process and will be valuable in designing a strategy for reducing the errors in energy forecasting. (c) 2005 Elsevier Ltd. All rights reserved.