This study examines whether household access to microfinance reduces poverty in Pakistan and, if so, how and to what extent. It draws on primary empirical data gathered by interviewing 1132 households, including both borrower and non-borrower households, in 2008–2009. Sample selection biases have been partially controlled for by using propensity score matching. The study reveals that microfinance programmes had a positive impact on the participating households. Poverty-reducing effects were observed on a number of indicators, including expenditure on healthcare, clothing and household income, and on certain dwelling characteristics, such as water supply and the quality of roofing and walls.