The prevailing paradigm attributing biotic mass extinctions to Large Igneous Province (LIP)-driven global hyperthermia and oceanic anoxia faces a fundamental sedimentological paradox: coal seams and carbonaceous shales — indicative of cool-temperate climates — are systematically developed in the strata of all five major Phanerozoic extinction events, while typical hot-climate indicators such as red beds and evaporites are conspicuously absent from those same intervals. This contradiction strongly suggests that the climatic backdrop of extinction intervals may be cold rather than hot. To test this possibility, we apply an a priori, falsifiable lithological criterion — the Cold Rule — that is independent of geochemical proxies. Taking coal and carbonaceous shales as cool-temperate indicators and red beds and evaporites as tropical indicators, we systematically compile published stratigraphic successions from key sections of all five mass extinctions, assign lithology to climate zones bed by bed, and examine three independent lines of evidence: climate reverse cycles, the timing of coal measures relative to extinction pulses, and the latitudinal differentiation of extinction. Each mass extinction event records a complete climate reverse cycle from red beds (warm stage) through evaporites (arid-hot stage) and coal and carbonaceous shales (cool-temperate stage) to cold-zone or glacial conditions at the extinction horizon, across all five events. The horizons of coal and carbonaceous shale development consistently predate the main extinction pulse, indicating that they represent the climatic prelude to extinction rather than synchronous products of a hyperthermal–anoxic crisis. Extinction intensity is highest at mid-latitudes, while low-latitude equatorial regions serve as relative biological refugia; extinction timing displays a systematic progression from high latitudes through mid-latitudes to low latitudes. On the basis of this primary sedimentological evidence, we propose the "Cooling-Driven Mass Extinction" hypothesis. Its causal chain operates through two timescales: in the short term, volcanic eruptions and/or bolide impacts trigger abrupt cooling through stratospheric aerosol forcing; in the long term, low temperatures suppress organic matter decomposition and enhance carbon burial efficiency, initiating a positive feedback in which cooling lowers atmospheric CO₂ and drives further cooling, ultimately contracting climate zones until the cold zone invades mid-to-low latitudes and triggers ecosystem collapse. This hypothesis accommodates all five mass extinctions within a single climate-driven framework. Its core predictions — that cool-temperate lithologies must predate the extinction horizon, that extinction timing must display a latitudinal progression from high to low latitudes, and that tropical lithological indicators must be systematically absent at extinction boundaries — have received preliminary empirical support in the data analysis presented here and are directly testable by future high-resolution stratigraphic work.