Key Findings -Changes in district level crop productivity among smallholder farmers have strong and positive lagged multi-year effects on the own-farm incomes of rural households in that district. -This impact is especially true for productivity changes among (a) the highest productivity farms in each district, and (b) smallholder farms cultivating >2 hectares. -There is also some evidence of a similar effect on total income, however this effect is not as robust. -Overall, the least robust set of results are between district-level crop productivity and off-farm household incomes, suggesting that some of the recent critiques of the small farm-led multiplier effect hypothesis mentioned earlier for the African context may be valid. -However, we do find tentative evidence (interpreted with caution due to their lack of significance in the robustness checks) that smaller farm productivity (<2 hectares) indirectly raises off-farm incomes.