The objective of this paper is to empirically analyze the dynamic relationship between the banking system and real economic activity in Peru. The analysis is based on the estimation of cointegrated vector autoregressions, the application of exogeneity tests and the identification of permanent and transitory shocks. Using annual data for the period 1965-2011 and three traditional indicators of financial intermediation associated with the banking sector, the following results were obtained: (i) there is a long-run relationship between the evolution of the banking system and real per capita GDP, (ii) real per capita GDP contributes to the prediction of banking system evolution, (iii) a permanent shock has more important effects on both real per capita GDP and banking system evolution than a transitory shock does, and (iv) fluctuations in real per capita GDP are mainly associated with permanent shocks at all horizons, while the short-run fluctuations in the evolution of the banking system are explained mainly by transitory shocks. The results suggest that real per capita GDP can be used by the financial regulator to assess prudential measures related to the evolution of the banking system; furthermore, they make evident that financial regulation must ensure that the level of banking intermediation is consistent with the evolution of real per capita income.