The finance-growth nexus is a classic source of debate among economists. This paper offers regional evidence on this issue in order to determine whether it can fit the data on a 147-year-old economic union, Italy. By means of this approach the pooling of developed and developing countries in the same sample can be avoided. Both cross-sectional and panel data estimates appear to show that more finance generates more growth. Endogeneity does not bias the results to a significant extent, and the finance-growth nexus is robust to spatial unobserved heterogeneity. Spatial correlation in the residuals is rejected by the data. Economic growth appears to be favoured more by short-term than by long-term credit.