We present a simple model which establishes a non linear and possibly non monotonic relationship between financial development and economic growth. Applying a threshold regression model to King and Levine™s (1993) data set, we find evidence that is consistent with the main implications stemming from the theoretical model.