There is hardly any argument over the necessity of targeting the ultra-poor in development interventions. However, identifying and scaling up effective strategies to improve livelihoods remains a challenge. A few recent pilots have found an approach that combines transfer of productive assets, and intensive supports and supervision with a set of coordinated interventions following a time-bound exit plan successful. This paper evaluates one such pilot, known as ‘ultra-poor graduation pilot’, implemented by BRAC in Afghanistan. We find that participating households have successfully utilised the livestock received from the programme to increase employment and income. This income growth has also allowed them to improve dietary diversity through consumption of more home-grown food items. Beneficiaries also increased their livestock holdings beyond what they received from the programme. We find relatively stronger impacts on investments compared to consumption indicating a shift towards longer-term planning among the beneficiaries. Given its effectiveness, this model should be strongly considered for scale-up in Afghanistan. We also suggest a couple of potential areas that need to be considered for improving the effectiveness of this model even further.