We study the relation between credit conditions, misallocation of resources, and total factor productivity (TFP) using sectoral data from Mexican manufacturing industries between 2003 and 2010. Our analysis uses a theory-based framework to account for TFP changes in the Mexican manufacturing sector due to changes in distortions in the use of capital, labor and intermediates arising from nancial frictions. We nd empirically that these distortions account for a large fraction of aggregate TFP changes in the period. We also show that changes in distortions in the data are statistically related to changes in the availability and the cost of credit. Taken together, the results suggest a connection between credit conditions and productivity at the sectoral level, channeled through the choice of the inputs mix by fi rms. Moreover, our analysis suggests that the reallocation of credit from distorted sectors to undistorted sectors is as important as the overall credit availability to explain TFP growth, especially during the recovery after the crisis.