Background and aim: Evidence on determinants of prices for orphan medicines is scarce and not available for Italy. The aim of this paper is to provide an evidence on variables affecting the annual treatment cost of orphan drugs in Italy, testing the hypothesis of a negative correlation with the dimension of the target population and a positive correlation with the added therapeutic value of the drug and the quality of the evidence of pivotal studies. Methods: Drugs with a European orphan designation reimbursed in Italy in the last 6 years (2014-2019) were considered. Univariate, cluster analysis and multiple regression models were used to investigate the correlation between the annual treatment cost and, as explanatory variables, the dimension of the target population, the existence of Randomized Clinical Trials as a proxy of the quality of the pivotal studies, the added therapeutic value. Results: In the univariate analysis prevalence and added therapeutic value, as expected, have a negative and positive correlation with cost respectively. The correlation with RCT is not significant. In the multivariate model, coefficients for prevalence and added value are confirmed but for the latter are not significant anymore. We also found, through an interaction analysis, that the existence of an RCT has a positive impact on annual treatment cost when the target population is very small. Conclusions: Our evidence suggests that value arguments and sustainability (dimension of the target population and its impact on budget impact) issues are considered for orphan drugs pricing: the role played by sustainability is systematically supported by the statistical evidence. A more transparent and reproducible price negotiation framework for orphan drugs is needed in Italy. This paper has contributed to highlight the implicit drivers of this framework.