In this paper we develop a switching regression model of investment, in which the probability of a firm facing a high premium on external finance is endogenously determined. This approach allows one to address the potential problem of static and dynamic misclassification encountered where firms are sorted using a criteria chosen a priori. We use U.S. firm level data to analyze the effects of variables that capture each firm's credit worthiness, asymmetric information, and agency problems on the probability of being in the high- or low-premium regime. The role of macroeconomic conditions and monetary policy is also discussed.