A forecasting model (BSPcast) developed for prediction of brown spot (Stemphylium vesicarium) of pear was evaluated as an advisory system for reduced fungicide use in disease management programs. Eleven trials were performed during 1995, 1996, and 1997 in five orchards in two different climatic areas in Catalunya (Spain) and Emilia-Romagna (Italy). Values of 3-day cumulative daily infection risk (CR) provided by the model were used to determine risk periods during the growing season of pear and were taken as thresholds to schedule fungicide sprays. The fungicide application programs tested using the model consisted of guided schedules with CR action thresholds of 0.4, 0.5, or 0.6, and fixed sprays following a standard commercial schedule. In nine out of 10 trials, no significant differences were observed in disease incidence on fruit at harvest between the fixed-spray commercial schedule and guided sprays using thresholds of 0.4 or 0.5. The average savings in number of fungicide sprays applied using BSPcast compared with the fixed-spray schedule were 20 to 70% when using fungicides with a 15-day protection period (kresoxim-methyl or procymidone) and ranged from 20 to 50% when using a fungicide with a 7-day protection period (thiram).