ABSTRACT. Native harvest statistics are counts, or estimates, of the number of animals by category taken by a specific group of native people during a specific time period. These statistics are significant for basic research in the social and biological sciences, for public policy and for the resolution on environmental conflicts in the North. This paper reviews and assesses two common sources of native harvest data- administrative and monitoring records, and special-purpose studies - and provides an extensive bibliography for the latter. Native harvest data are normally obtained by recall survey rather than direct observation. The existing data base is therefore evaluated in terms of the methodological norms of social surveys, with particular attention to precision and uniformity of survey parameters and interview terminology, sampling procedures, non-response bias and response bias. Despite some lack of methodological rigour, especially regarding parameters, terminology and projection from reported harvest, it is, concluded that the existing body of information may be used to recreate an historical statistical series of substantial breadth and depth, useful for both socio-economic and biological research purposes.