This article analyses the effects of public expenditure on the Bolsa Família Program on the Brazilian crime rate. It covers 10 of the 12 years of the Program's intervention and suggests a unique approach in the national literature. The empirical strategy is based on a balanced regression panel model with fixed effects, using data from the 27 Brazilian Federative units between 2005 and 2014, totalizing 270 observations and 27 cross sections. The dependent variable of the model is the number of intentional homicides per 100.000 inhabitants. The difference of this article to other empirical studies is the inclusion, as an independent variable of the model, of the public expenditure on the Bolsa Família. The results suggest a negative relationship between homicide rates and the expenditure on the Bolsa Família and indicate that the Program may have extensive social consequences to be considered when social policy is designed. Therefore, the results of this article become important to support future Program adjustments and suggest that Conditional Welfare Programs may reduce violent crimes in developing countries.