Accountability is a major challenge for the Farming Systems Research/Extension approach. After its enthusiastic embrace in the 1970s and 80s, calls have gone out to provide evidence to justify the large investments that have been made. This paper presents the results of a study carried out in the middle-belt region of Nigeria to evaluate the impact of an FSR/E project implemented by the International Institute of Tropical Agriculture on the adoption of a package of technology for lowland rice by resource-poor farm-households. FSR/E participants, in comparison to non-participants, showed statistically significant higher adoption rates. Participation in FSR/E was also the single most powerful predictor of adoption, while classical diffusion variables were poor predictors. It is recommended that rather than discard the FSR/E model, steps should be taken to strengthen farmers’ participation and to make the model more responsive to the needs of resource-poor farmers.