Long-term changes in the fish assemblage of the Big Blue River basin, Kansas, were examined by comparison of a stream survey conducted prior to the construction of Tuttle Creek Reservoir (1957 to 1958) with two surveys conducted approximately 40 years later. The most recent survey, in 2001, included 14 of the original sites sampled in the late 1950s. The most notable changes to the fish assemblages were the introduction and presumed establishment of four species (Notropis atherinoides, Pimephales vigilax, Gambusia affinis, and Notemigonus crysoleucas) and the apparent extirpation of Macrhybopsis aestivalis. Other than the loss of M. aestivalis, incidence and relative abundance of other species remained similar to that which occurred before the construction of the reservoir. Percent similarity of the fish assemblage between the 1950s and the recent samples ranged between 34.7% and 84.4%, dependent on stream type (mud, gravel, or sand substrate). In addition, species incidence patterns were significantly concordant between the 1950s and the two recent surveys, with the exception of mud-bottomed streams sampled during 2001. Finally, Jaccard's index was used to compare assemblage similarity between 14 sites sampled in both the late 1950s and 2001. Although assemblage similarity between time periods was on average low (x̄ = 0.411), a Mantel test suggested a significant concordance between sampling periods, indicating the spatial structure of the assemblage was stable. We conclude that Tuttle Creek Reservoir has served as a source for the establishment and subsequent dispersal of introduced species in the Big Blue River system and that it is important to monitor routinely these assemblages because species may continue to invade the Big Blue River system with unpredictable consequences for the native fish fauna.