The current status and future prospects of long-range weather forecasting are reviewed. Recent climate-related disasters, and the prospect of major climate change in the coming decades, have increased scientific and public interest in the causes and predictability of climate variation. Some new long-range forecast systems have been introduced recently. Older systems have been improved by changes in their presentation and dissemination as well as by improvements in our knowledge of the causes of climate fluctuation. New methods for forecasting have been proposed including global sea surface temperature patterns, and links between sunspot cycles, the quasi-biennial oscillation and the atmosphere. Substantial improvements in our understanding of the El Nino Southern Oscillation phenomenon, and its potential use in forecasting, have been made. Many of the new forecast systems rely heavily on the predictability of this phenomenon. The increased interest in the subject has led to greater awareness of some of the dangers inherent in developing long-range forecast systems based on limited data and a large number of potential predictors. It has also resulted in the recognition that conventional long-range forecast products, monthly and seasonal rainfall totals and mean temperature, may not be the most appropriate. andrdquo;Designerandrdquo; forecasts which attempt to match the abilities of forecasters with the needs of potential forecast users, seem likely to appear in the near future. Overall, the prospects for improved long-range forecasting, particularly for the semiarid tropics and subtropics, appear promising.