Criticality assessments of raw materials are inherently based on multiple criteria, which justifies the use of multi-criteria decision analysis (MCDA) to aid the interpretation of the data by providing a comprehensive evaluation. A structured and transparent selection procedure is firstly introduced in this paper to choose eight supply risk assessment criteria to evaluate the security of supply for thirty-one raw materials used in automotive manufacturing. A synergic combination of MCDA methods is then proposed for the classification of raw materials in risk classes according to the supply risk criteria. Risk classes are recommended following from a robustness analysis based on stochastic and optimisation MCDA methods where risk levels assigned to the raw materials are firstly visualised on a relative frequency basis. The sorting of the raw materials is also refined by narrowing down the best and worst plausible classes when justifiable constraints on criteria weights are accounted for in the modeling. For example, the robustness analysis suggests that rare earth elements and tellurium have a high eventuality of supply chain disruption, closely followed by indium, germanium and boron. Conversely, the results suggest that the risk of supply disruption for iron, copper, zinc and aluminium is mostly medium-low or low. The proposed step-wise decision support approach can be used as a complementary tool to the existing life cycle assessment methods for a more comprehensive assessment of the short-term availability of natural resources.