In this paper we analyze the effects of some distortions induced by prospective payment system, i.e. Upcoding, Cream Skimming and Readmissions on hospitals' technical efficiency. We estimate a production function using a population-based dataset composed by all active hospitals in an Italian region, Lombardy, during the period 1998-2007. We show that cream skimming and upcoding have a negative impact on hospitals' technical efficiency, while readmissions have a positive effect. Moreover, we find that private hospitals are more engaged in cream skimming than public and not-for-profit ones, while we observe no ownership differences regarding upcoding. Not-for-profit hospitals have the highest readmission index. Last, not-for-profit and public hospitals have the same efficiency levels, while private hospitals have the lowest technical efficiency.