In past investigations the pattern of differential survival of plants across the K/Pg boundary has been viewed as incompatible with severe asteroid impact winter scenarios (i.e., an impact winter lasting more than a few months), particularly the enigmatic survival of coryphoid palms and Pandanus (screw pine). Stateof- the-art climate models based on soot, sulfate and nano-sized dust aerosols predict a global impact winter that drastically reduced precipitation and resulted in a transient period of total darkness and permafrost conditions. This suggests that the plants most likely to have been affected by the global mass-extinction event were tropical phanerophytes that produce recalcitrant seeds, which by definition are desiccation-intolerant, survive less than a year, and cannot survive freezing. However, this hypothesis has never been tested. In this study I sampled over 100 plant species from the global fossil record that have a high probability of having produced either recalcitrant seeds/disseminules (n1 = 58) or orthodox seeds (n2 = 59), based on their phylogenetic relationships with extant taxa that either are monomorphic for these traits or specifically exhibit a genetic marker for abscisic acid inhibition associated with seed dormancy and recalcitrance. A one-tailed z-test for the difference between two proportions revealed that plant taxa with a high probability of having produced recalcitrant seeds had significantly lower survivorship than plant taxa with a high probability of having produced orthodox seeds (p < 0.0001). Based on these data, it can be concluded that plants which formed a frost-tolerant seed bank during the latest Maastrichtian were significantly more likely to survive the K/Pg impact winter than plants which did not (including palms). These data clearly indicate that the K/Pg impact winter probably lasted longer than a year and that it selected for seed-based traits that effectively sorted correlated functional traits of mature plants (i.e., leaf physiognomic features). This novel hypothesis stands as an alternative to J.A. Wolfe’s classic hypothesis that a mild K/Pg impact winter selected for fast-growing angiosperms with deciduous leaves and did not affect the plant communities of the Southern Hemisphere. Potential mechanisms for the rare survival of tropical, recalcitrant-seeded plants are discussed.