We organise a field experiment with smallholder farmers in Rwanda to measure the impact of financial literacy training on financial knowledge and behaviour. The training increased financial literacy of participants, changed their savings and borrowing behaviour and had a positive effect on the new business start-up. However, it failed to have a significant (short-term) impact on income. Using a two-stage regression framework, we identify enhanced financial literacy as one of the important factors explaining behavioural changes. We also test whether financial knowledge spillovers from trained farmers to their peers in local village banks but find no evidence for that.