Anadromous fish, and salmon in particular, have frequently been implicated in archaeological models of social complexity for hunter-gatherers. It has been assumed that Atlantic salmon was once an important resource, prior to its extinction in the rivers of southern New England in the eighteenth century, to northeastern prehistoric aboriginal peoples. This perceived importance stems from referents to Atlantic salmon in various historical accounts. These accounts have been interpreted, especially by fisheries biologists involved with salmon restoration programs, and consequently by anthropologists, to indicate a past resource that rivaled present Pacific salmon or Atlantic cod in relative biomass. Archaeological evidence in the form of fish bones recovered from seventy-five sites in the northeastern region, dating from Middle Archaic to Woodland periods (8,000 B.P. to Contact), indicate the virtual absence of Atlantic salmon in the prehistoric period while other types of fish are preserved. This is supported by a complete absence of salmon in the paleontological record for the late glacial/early postglacial period. This research evaluates various cultural and biological possibilities to explain the virtual lack of salmon archaeologically, and the discrepancy between that and the perceived multitudes of Atlantic salmon historically. Factors such as sampling bias, preservation, and lack of aboriginal salmon technology are discounted. Instead, the working hypothesis that salmon did not begin to colonize New England streams in substantial numbers until the historic period, corresponding to a more favorable period of climatic cooling known as the Little Ice Age (A.D. 1550-1800), is presented. Environmental and climatic factors affecting range shifts in salmon and other fish are investigated, as well as genetic relationships between European and American salmon stocks. The presumed abundance of Atlantic salmon based on historical records is also re-examined through a critical source analysis of the social and environmental contexts of the texts, concluding that the great abundance of salmon has been exaggerated both by early and recent writers. The implications for archaeological reconstructions of prehistoric New England aboriginal lifeways without a rich salmon resource are discussed and compared with models of surplus production, storage, and social complexity, for aboriginal peoples inhabiting the Northwest Coast Pacific salmon region. The research concludes that these models are inappropriate for understanding prehistoric cultural developments in New England. In addition, the possibility of a fundamental environmental basis for salmon range expansion and subsequent retraction associated with the beginning and end of the Little Ice Age, as opposed to an anthropogenic basis of pollution and dam building, may have implications for the limited success of biological salmon restoration programs.