Several outbreaks of forest pathogens have occurred in North America in historic time. The chestnut blight (1904-1950), which virtually exterminated Castanea dentata, has left a clear fossil record. In subsurface sediment from Connecticut lakes chestnut pollen drops from 7% to less than 1% and just above these levels first birch pollen (probably from Betula lenta, a successional species) and then pollen from oak (a com· petitor of chestnut) increase in abundance (Brugam, 1975). A decline in chestnut pollen can be seen everywhere where chestnut was abundant in the forest. A similar phenomenon occurs in sediment deposited 4800 radiocarbon years ago. Tsuga pollen declined precipitously throughout the geographical range occupied by the species at that time. In New Hampshire it dropped from 30% to 5% within a century. In some regions it reg\ined its previous abundance 2000 years later, but in other regions the forest community changed through immigration of new species, and hemlock never recovered full abundance. The synchroneity of the hemlock fall at 4800 radiocarbon years, at. sites from New Brunswick to Upper Michigan (a distance of 1500 km) despite different abundances of hemlock and different forest communities is persuasive evidence for the widespread outbreak of a pathogen on hemlock. The swiftness of the decline and its widespread occurrence argue against climatic change as the causal factor. Neither could human activity have been important simultaneously over such a wide region. In each locality, a series of changes in tree pollen abundances follow the hemlock decline; these reflect a successional sequence that differed in each forest community. Although disease may have played a part in hastening the local demise of other tree species at times when climate changed, there is no evidence elsewhere in the Holocene pollen record for an outbreak of a paihogen. The elm decline in Europe 5000 years ago is the sole exception but it is often attributed to climate, or to human influence because it was simultaneous with the advent of agriculture. The rarity of sudden sharp declines in pollen abundances over wide regions implies that pathogen outbreaks have been infrequent phenomena in Quaternary history.