Computer software for forecasting rice blast epidemics has undergone remarkable changes with the advances in computer hardware in the last decade inJapan. In the early stage of that decade, several predicting equations were proposed by either multiple regression or quantification analysis. A system for estimating favorable conditions for infection based on empirical criteria was developed. The weather data used in the system were retrieved from the Automated Meteorological Data Acquisition System by personal computer via telephone lines. Several simulators incorporating ecological knowledge were subsequently developed. For example, the simulator of leaf blast epidemics, a deterministic model, was originally designed for a mainframe computer, and was later improved to be easily executed on a 16 bit personal computer. It can also estimate the effectiveness of some fungicides in the progression of the disease. Another simulator dealing with panicle blast epidemics, is a stochastic model using the Monte Carlo method which is able to simulate the disease progression within a panicle and to estimate the yield losses caused by panicle blast. However, temporary values are substituted for several processes which have not yet been verified experimentally.