International agricultural research often uses quasi-experimental designs when implementing on-farm research and development activities. Therefore, impact assessment methodologies are needed, which are effective in circumventing the resulting selection bias inherent in such research designs. This paper applies propensity score matching (PSM) as one way to measure the impact of ILRI-led research activities to control African animal trypanosomosis (AAT) in West Africa. Data were collected from 508 farmers in Mali and Burkina Faso. Results indicate significant improvements in farmers’ knowledge. The paper adds to the methodology of PSM in impact assessment by emphasizing on the quality of different matching algorithms and on the sensitivity of impact estimates.