OBJECTIVES: The STERRAD® Sterilizers are a versatile group of systems that use low-temperature, hydrogen peroxide gas plasma technology to sterilize medical devices safely and effectively. Literature data suggest significant reductions in the number of damaged rigid endoscopes with STERRAD® compared to steam sterilization. The aim of this study was to assess the potential savings associated with the STERRAD®100NX™ system compared to steam for endoscope sterilization. METHODS: A dynamic budget-impact model was developed to estimate the costs with the use of the STERRAD® 100NX™ system compared to steam, from a hospital decision-maker perspective and with a one-year time-horizon. Data were collected from two Italian hospitals running approximately 800 sterilization cycles per year, and costs of sterilization per cycle, endoscope repair costs, testing and maintenance costs were considered. A conservative estimate for the endoscope repair cost reduction of 33% was used in the base case analysis. Gamma distributions for cost parameters and Bootstrap methods were used to generate 1000 simulations. The acquisition costs of STERRAD®100NX™ and steam systems were not considered. RESULTS: In the base case analysis, the annual costs with STERRAD® were €55,540, and the mean costs (SD) from the simulations were €51,471 (€25,186). For steam sterilization, the corresponding figures were annual costs of €59,736 for the base case, resulting in a cost-saving with STERRAD® of €4,196, and a mean cost of €58,078 (SD €33,785) from the simulations, giving mean probabilistic cost-savings with STERRAD® of €6,608. With a reduction in endoscope repair costs of 41.6%, as calculated from published studies, savings with STERRAD®increased to €8,803. CONCLUSIONS: This analysis suggests that the use of STERRAD® for the sterilization of endoscopes may result in important cost-savings compared to reprocessing with steam. Costs for the sterilization of other reusable heat-sensitive devices were not considered; therefore these cost-savings may have been underestimated.