The Farmer Field School (FFS) is an intensive training program providing farmers with science-based knowledge and practices, including integrated pest management. Recently there has been intensive debate as to whether or not this kind of training has any significant impact. Most case studies argue that the impact, in terms of a farmer's ability to reduce the use of pesticides while increasing yields, is significant. However, panel data studies using household panel data sets for Indonesia have not been able to confirm that this is the case. The present paper uses panel data available from previous panel data studies and applies a new model specification to reevaluate whether or not the FFS induces better performances among farmers enrolled in the program and also among their neighbors, who are expected to receive some spillover knowledge from the FFS alumna.