Fish populations were sampled in two quarter-mile sections of the Coosa River near Childersburg, Alabama, during July and August, 1949. Sampling was done with hoop nets, wire baskets, and rotenone. In addition, the catch of a permanent fish trap was observed. The selectivity and limitations of each of these sampling methods are discussed. Since the methods used were to an appreciable extent complementary, it was possible to obtain significant information on nearly all important species. The most important species (by weight) taken were channel catfish, freshwater drum, smallmouth buffalo, flathead catfish, blue catfish, redhorse, and gizzard shad. None of the methods used was adequate for the sampling of game and pan species. The most important of these encountered were, however, white crappie, spotted black bass, and bluegill sunfish. Each of the methods failed completely to reveal the presence of one or more important species of fish. Since all sampling methods were selective, it was impossible to arrive at estimates of the relative or absolute abundance of the various species of fish. However, figures are given on the relative and absolute availability of the different species to the various sampling methods. The success of mark-and-recapture attempts to obtain estimates of relative abundance which were independent of the species selectivity of the gear was limited by the very small number of recaptures.