Hospital heterogeneity is a major issue in defining a reimbursement system. If hospitals are heterogeneous, it is difficult to distinguish which part of the differences in costs is due to cost containment efforts and which part cannot be reduced, because it is due to other unobserved sources of hospital heterogeneity. In this paper, we apply an econometric approach to analyse hospital cost variability. We use a nested three-dimensional database (stays-hospitals-years) in order to explore the sources of variation in hospital costs, taking into account unobservable components of hospital cost heterogeneity. The three-dimensional structure of our data makes it possible to identify transitory and permanent components of hospital cost heterogeneity. Econometric estimates are performed on a sample of 7314 stays for acute myocardial infarction (AMI) observed in 36 French public hospitals over the period 1994-1997. Transitory unobservable hospital heterogeneity is far from negligible: its estimated standard error is about 50% of the standard error we estimate for cost variability due to permanent unobservable heterogeneity between hospitals. [Ed.]