This paper tackles one of the most complex and multifaceted agricultural policy decisions faced by African governments; that is, whether it makes sense to subsidise fertilisers. It combines a data envelopment analysis to compute efficiency scores and regression analyses that deal with self‐selectivity and endogeneity issues to examine the productive efficiency of farmers who benefited from the subsidy programme in the very atypical irrigated system of the Senegal River Valley. The results based on farm‐level data indicate that the subsidy programme seems to be working, as it appears to be associated with increased efficiency; hence, providing empirical support to the political will to revamp subsidy programmes. The results also suggest various ways to improve the effectiveness of the subsidy programme in conjunction with additional policy options to further unlock the agricultural potentials of the region.