The paper focuses on Supply Chain Risk (SCR) early-warning system design. Based on artificial immunity principle, a SCR early-warning conceptual model is established first. A SCR early-warning system is presented with detailed early-warning indexes and risk grading. The early-warning method is shown by a practical case. Applying the early-warning system to make cluster analysis for SCR grading, it provides a new perspective for study and application of SCR early-warning.