INTRODUCTION: Road safety is a recognized global issue and according to the WHO, road traffic injuries are the eighth leading cause of death in all age groups, especially 5 to 29 years. Therefore, in this article, the road safety performance of Iran's provinces is examined. METHODS: This research was done using Data Envelopment Analysis (DEA) method which is used in two deterministic and non-deterministic situations in order to evaluate road safety efficiency scores. This method gives scores (inefficiency) that allow road sections to be ranked appropriately in terms of being accident-prone. Uncertainty is one of the inevitable features of real-world problems, for which fuzzy theory and extend the DEA-RS model is used by considering its limitations as probability, necessity, and credibility constraints, and propose three fuzzy models such as possibility of DEA-RS (PosDEA-RS); necessary DEA-RS (NecDEA-RS); and the credibility of the DEA-RS (CreDEA-RS). FINDINGS: Three models which are extensions of the Data Envelopment Analysis based on the Road Safety (DEA-RS) model are proposed for evaluating road safety performance and the CreDEA-RS model is suitable for assessing the safety of roads in the provinces of Iran. CONCLUSION: The results show that the provinces located in mountain and forest areas like Gilan have a lower performance in terms of road safety, and provinces located in desert areas like Yazd have a higher road safety performance.
