We conduct a field experiment in which we offer graduated microfinance borrowers a much larger loan, to purchase a fixed business asset. We implement this using a shared-ownership hire-purchase contract; our control group receives a zero-interest microcredit contract. We find large, significant and persistent effects from the hire-purchase contracts relative to the zero-interest microcredit: treated microenterprise owners are more likely to remain in self-employment, run larger businesses, and enjoy higher profits. As a result, their households consume more —- particularly including increased spending on food and on children’s education. We test for heterogeneous effects, and find that a flexible repayment schedule is relatively more attractive for risk averse clients. To understand likely mechanisms driving our results, we estimate a dynamic structural model; this model emphasises capital lumpiness, and implies an important role for large capital purchases in shifting enterprise trajectories.