Maternal tolerance of the semi-allogeneic fetus is fundamental to successful pregnancy, yet traditional “immune suppression” models fail to explain three key clinical phenomena: the marked heterogeneity in pregnancy outcomes, the abrupt and refractory onset of preeclampsia, and the increased incidence of autoimmune diseases in the postpartum period. Here we propose a conceptual theoretical framework—the “inflammatory load-driven bistable switch model of maternal–fetal immune tolerance”—in which the maternal–fetal immune system transitions between two stable states: a tolerance state and an effector/inflammatory state. Transition between these states is governed by the interplay between inflammatory load (cumulative pressure from infection, hypoxia, tissue damage, etc.) and the tolerance reserve index (S-value). As pregnancy advances, sustained fetal antigen exposure may progressively promote the emergence of PD-1⁺TCF-1⁺ precursor exhausted-like states, gradually reducing the S-value. When inflammatory load exceeds the critical threshold of S, the system undergoes an irreversible state transition—programmed and controlled at term as normal labor, or premature and excessive as preeclampsia. Postpartum, delayed Treg reconstitution relative to effector T cell recovery creates an immune imbalance window predisposing to autoimmune diseases. This framework provides a unified mechanistic explanation for diverse pregnancy complications and generates testable clinical predictions for prospective validation.