n this article, we evaluate the efficiency of the 10 different regions of Ghana using slack-based data envelopment analysis, a nonparametric linear programming technique. Furthermore, we analyse the variable effects on the efficiency of the regions by various regression models using bootstrap sampling technique. The data come from the 1991/1992 and 1998/1999 Ghana Living Standards Survey. Our results show that wealth is not strongly related to efficiency. For example, the study indicates that the Brong–Ahafo region is the most efficient region but not the most wealthy in Ghana. Generally, urban regions are not found to be among the most efficient regions due to the high expenditures. The regression analysis shows that female heads of household have an overall positive effect on efficiency. In addition, any form of education obtained is also found to have a significant positive effect on efficiency.