We use a combination of detailed panel data and a randomized experiment to assess the role of capital in the recovery process of microenterprises in Sri Lanka following the December 2004 tsunami. Disaster relief in low-income countries typically comes in the form of relief aid rather than insurance payments. In our data aid is found to be uncorrelated with the extent of business losses and to cover only a small fraction of losses. Our results show that lack of access to capital inhibits the recovery process, with firms receiving randomly given grants recovering profit levels two or more years before other damaged firms. Access to capital appears to be particularly important for the retail sector, whereas reductions in demand, disruptions in supply chains, and other disasterinduced changes appear to limit the extent to which capital alone can spur recovery in the manufacturing and services sectors. Overall our data show that business recovery is much slower than commonly assumed, providing a role for targeted aid to hasten microenterprise recovery dramatically following such disasters. JEL Classification Codes: O12, Q54, C93.