The purpose of this study is to analyze the impact of PRONAF credit program on small-farm agricultural production in Brazil. The study compares farmers’ production value considering the obtainment of PRONAF credit, controlling for farm, farms and production system characteristics. The data set consists of the 2006 Agricultural Census, which considers 5.2 million of small farmers in Brazil. In addition to using multiple linear regression model to estimate the net impact of PRONAF on total production value, we applied a propensity score matching method in order to identify pairs of family farms relatively homogeneous, one that accessed the credit and other that did not, estimating the average difference between their production values. Regression analysis showed that the access to PRONAF had a positive and significant net effect on production value of around 18%. In addition, propensity score matching results seemed to exhibit similar evidence to those obtained by regression model. Farmers that obtained PRONAF microcredit presented a production value higher than others, with the difference ranging from 6% to 20%. The impact is lower in the less developed regions, which is characterized by forestry, subsistence agriculture and low technology adoption. For more developed regions, where farmers are more specialized and integrated in the market, the PRONAF has shown relevant net impacts on the production value.