According to this article, reliable advance information about likely meteorological conditions during an agricultural season potentially greatly benefits planning, risk management and productivity. This article reviews the present state of production and dissemination of meteorological long-range (seasonal) forecast information and the use of such products in agricultural applications. There has been rapid development of dynamical prediction systems in particular, and several centres routinely provide seasonal forecasts of rainfall and temperature with global coverage. Currently the uptake of long-range forecast products by users has been limited, and the potential value is far from being attained. Various factors that inhibit usage are described. Further investment in the infrastructure is required, both in creating relevant specific products and in disseminating and applying them effectively. Investment in research is also required to investigate opportunities and beneficial strategies for a wide range of regions and activities, with closer interaction between the meteorological and agricultural communities and relevant intermediary agencies. Such development would complement separate ongoing efforts to improve the meteorological forecast systems and to improve agricultural management and technology.