A critical issue surrounding biomonitoring in large rivers (fifth- through seventh-order) is the minimum sampling-reach distance required to collect an adequate number of fish to represent the fish assemblage within a reach. Excessive sampling effort (excessive reach length) is costly in terms of work hours, reduces the number of sites that can be visited, can compromise field-crew safety, can be logistically unfeasible, and can cause unnecessary injury to captured fish. On the other hand, inadequate sampling effort can produce considerable variability in multiple samples collected at a site and may underrepresent the species or river condition present. During the summer of 2002, the U.S. Geological Survey, in cooperation with the Idaho Department of Environmental Quality, determined the minimum sampling effort required to characterize fish assemblages at 17 large-river sites in southern Idaho. The study was done as part of the U.S. Environmental Protection Agency's Environmental Monitoring and Assessment Program. Electrofishing methods and multiple gear types were used to collect sample populations of fish in river reach lengths representing 40 and 100 times the wetted channel width. Minimum sampling effort was assessed by comparing the relation between reach length and the number of species collected, total individuals collected, and final Index of Biotic Integrity (IBI) scores. Thirty-two species of fish in the families Catostomidae, Centrarchidae, Cottidae, Cyprinidae, Ictaluridae, Percidae, and Salmonidae were collected. Of these, 12 alien species were collected, representing about 38 percent of all species collected during the study. Wetted channel width was determined to be sufficient for collecting an adequate number of fish to estimate species richness and evaluate biotic integrity. At most sites, about 250 fish were needed to effectively represent 95 percent of the species present. Fifty-three percent of the sites assessed, using an IBI developed specifically for large Idaho rivers, received scores of less than 50, indicating poor biotic integrity.