This study introduces reverse stress testing for supply chains, designed to identify the minimal deviations from normal operations that would drive supply chains to a predefined performance failure. First, we present a framework for reverse stress testing with the purpose of assessing supply chain vulnerabilities. The framework involves six steps: selecting risk variables, defining baselines, formalizing the policy, formalizing key performance indicators (KPIs), determining critical thresholds, and identifying the critical scenario. Building on this framework, we propose an optimization model that identifies the critical scenario by uncovering the joint risk variable deviations that stress the supply chain beyond critical KPI thresholds. We also show that the optimization model has desirable theoretical properties, demonstrating that the model is computationally tractable, that a stress scenario always exists under mild assumptions, and that its output establishes a robustness guarantee for the policy. Two numerical examples demonstrate the frameworkâs applicability: a serial supply chain model and a service guarantee model. Results show that the framework effectively uncovers where disruptions become most harmful.