The purpose of this paper is to identify the non-lineal and asymmetric behavior of the banking credit for the case of Peru during 1994 - 2010. These characteristics are important because monetary policy, financial regulation, and business strategies from the part of the banks change according to the state of the economy. To this purpose, I use two-regime models: LSTAR and Markov Switching. Both of them identify the contraction state during 1999-2004. For these years, the probability of transition in between regimes may be related to the recessive impact of international crises. These results show that the expansionary regime is harsher, the series does not have a long memory, and its adjustment to different type of shocks is relatively faster.