This interpretation and compendium of historical fire accounts in the northern Great Plains provides resource managers with background information to justify the study or use of fire in management and provides a reference of historic fire accounts for those without ready access to major library collections. Historical accounts of fire are critiqued to aid interpreting the compendium accounts. An interpretation is included by the author.Lightning-set fires were recorded in the literature far less frequently than were Indian-set fires. The kinds of fire most frequently reported were scattered, single events of short duration and small extent. Although fires occurred in wetlands, wetlands as well as sandy soil sites usually were good areas for escape from the effects of fire. Both Indians and wild animals were reportedly injured or killed during prairie fires. The frequency of historic fires was less evident in the literature than the descriptions of fire distribution in time and space. Indian-set fires were reported in every month except January. Fires occurred mainly in two periods, March through May with a peak in April, and July to early November with a peak in October. Grassland fuels burned readily within a few hours or days after rain and even during light snowfall.I agree with arguments that support the concept that Indians of the northern Great Plains generally did not subscribe to annual wholesale or promiscuous burning practices, but that they did purposely use fire as a tool to aid hunting and gathering of food and materials. Apparently, the northern plains Indians did not pattern their use of fire with the seasonal patterns of lightning fires. More likely they developed seasonal patterns of burning the prairies in harmony with bison (Bison bison) herd movements because the hunter-gatherer economy of these nomadic tribes was centrally focused and largely dependent on bison and bison ecology.