This paper explores theoretically and empirically the link between Financial Liberalization (FL) and the banking crisis that often follow. We also investigate the proposition, classical in development economics, that FL should result in an increase in supply of funds to the real sector. To accomplish this we first develop a theoretical model of a banking firm that operates under financial repression and is then subject to FL. The model yields the result that following FL there is an unambiguous increase in risk to the banking firm which implies a higher probability of a banking crisis following FL. Less formally, we also conclude that the presence of a explicit or implict deposit insurance scheme is likely to accentuate the incentives to engage in risk and the risk structure of the banking system. Moral hazard plays an important role in this increase in risk to the banking sector. This questions the ''innocence'' of the bank owners in the crisis that have often followed FL and that had been attributed to either macroeconomic policy, concomitant structural changes in the economy or left-over distortions from the financial repression period. The sign of the change in supply of credit to the real sector, however, is ambiguous. Then we test empirically the propositions resulting from this model using data of 73 banks (some of which may have become technically insolvent) from Greece, Malaysia, Mexico, Taiwan and Thailand. The tests tend to support the conclusions of the theoretical model, i.e. unambiguous increase in risk and, for the sample used, an unambiguous fall in loan supply as a proportion of funds available. Finally we draw policy implication with respect to bank supervision, forbearance and bank failure resolution procedures during the transition period, and about the so-called ''liberalization sequencing.''