This brief describes key findings from a rigorous seven-year evaluation of the first of these livelihood programmes, BRAC’s ‘Targeting the Ultra-Poor’ programme in rural Bangladesh. Targeted households increase earnings by 37% and improve their consumption, savings, and asset accumulation. Results from Bangladesh are consistent with evidence from randomised evaluations of pilots in six other countries, which suggest BRAC’s approach improves outcomes for the ultra-poor across diverse contexts. Longer-term evaluations (four and seven years later) suggest long-run impacts may be even larger than two-year effects.