Over the last 35 years, microfinance has been generally regarded as an effective policy tool in the fight against poverty. Yet, the question of whether access to credit leads to poverty reduction and improved wellbeing remains open. To address this question, we conduct a systematic review of the quantitative literature of microfinance's impacts in the developing world, and develop a theory of change that links inputs to impacts on several welfare outcomes. Overall, we find that the limited comparability of outcomes and the heterogeneity of microfinance-lending technologies, together with a considerable variation in socio-economic conditions and contexts in which impact studies have been conducted, render the interpretation and generalization of findings intricate. Our results indicate that, at best, microfinance induces short-term dynamism in the financial life of the poor; however, we do not find compelling evidence that this dynamism leads to increases in income, consumption, human capital and assets, and, ultimately, a reduction in poverty.