Do capital grants improve microenterprise productivity? We use the lens of a production function to re-examine two previous randomised controlled trials that allocated capital to microenterprises. We find that productivity is higher for treated firms, and accounts for about 20-30 percent of the revenue effects of capital grants. Although long-run estimates are noisy, point estimates indicate that these productivity effects are sustained six years after the grants. We explore possible mechanisms for this finding, and show that treatment tilts the asset composition towards durables with a higher technology component: a result consistent with a important role for capitalembodied technology. Mediation analysis confirms that virtually all of the effect of treatment on productivity can be explained by the adoption of higher-technology durables.