Inventory management in the automotive industry has strategic importance due to the high variety of parts, the strong interdependence of production processes, and the high sensitivity to production line downtime. The traditional ABC classification, as one of the common inventory management tools, only considers the annual consumption value and is unable to realistically reflect the complexities, risks, and uncertainties in the supply chain. Therefore, the main objective of this research is to develop an intuitionistic fuzzy MADM framework to improve the classification of inventory items in the automotive industry. In this study a hybrid approach of ABC inventory classification and intuitionistic fuzzy MADM has been developed. Initially, the criteria of inventory classification have been determined by literature review and consulting experts, and then the criteria weights have been determined by intuitionistic fuzzy DEMATEL. Then inventory item ranking has been done by intuitionistic fuzzy EDAS. Finally, using the scores obtained from IF-EDAS, the inventory items have been assigned to classes A, B and C. The results show that the criteria of inventory classification are: item value, demand, lead time, shortage impact, supply risk, functional importance, part feeding method and inventory security sensitivity. The most important criteria are lead time, supply risk and inventory security sensitivity. Also the most important and sensitive inventory items are items related to the suspension, steering, and braking systems, engine and power transmission parts and electrical and electronic parts.