We review the empirical evidence on the impact of microcredit on poverty in Bangladesh. Drawing on evidence from seven empirical studies with 306 estimates, we examine the impact of microcredit on three proxies of poverty – income, assets and consumption/expenditure. After addressing issues of publication selection bias, we find that microcredit has a statistically insignificant effect on income, and also on assets. Evidence shows a positive but weak effect of microcredit on consumption/expenditure. Meta-regression analysis reveals that sources of variations in the existing literature such as study design, data characteristics and empirical methodology can explain the differences in reported estimates.