AbstractIn 1979, Biesinger et al. described a technique for spatial and temporal matching of records of stream temperatures and fish sampling events to obtain estimates of yearly temperature regimes for freshwater fishes of the United States. This article describes the state of this Fish and Temperature Database Matching System (FTDMS), its usage to estimate thermal requirements for fishes, some proposed maximum temperature tolerances for several freshwater fish species, and the way these FTDMS-derived values relate to various laboratory test results. Although applicable to all species for which collection records exist, initial development and refinement of FTDMS has focused on estimating the maximum weekly mean temperature tolerance for 30 common fishes of the United States. The method involves extensive use of automated data processing during data incorporation, quality assurance checks, data matching, and endpoint calculation. Maximum weekly mean temperatures derived from FTDMS were always less than laboratory-determined lethal temperatures and were similar to temperature criteria obtained from laboratory data through Environmental Protection Agency (EPA) interpolation procedures. The technique is a cost-effective means of generating temperature tolerance estimates for many U.S. fish species (i.e., more than 100).