Repopulation after a fish kill caused by pesticide pollution in Mud Creek, Washington County, Arkansas is reported. Twenty-nine species of fishes were eliminated from the study area as a result of the poison. Repopulation began almost immediately after the pesticide dissipated, and was accomplished primarily by immature individuals. Chronology and rate of repopulation of each species are reported, and factors influencing the rate of repopulation and population structure are discussed. With the number of fish kills increasing annually, the need for comprehensive understanding of fish repopulation is apparent. Gunning and Berra (1968) and Berra and Gunning (1970) investigated repopulation by sharpfin chubsuckers and longear sunfish, respectively, from experimentally decimated Louisiana streams. Repopulation after drought has been reported by Larimore, Childers, and Heckrotte (1959) in Illinois, Deacon and Metcalf (1961) in Kansas, and Harrel, Davis, and Dorris (1957) in Oklahoma. Krumholz and Minckley (1964) studied fish repopulation in a section of the Ohio River, and Cairns et al. (1971) discussed the biological damage and recovery from an ethyl benzene-creosote spill on the Roanoke River in Virginia. These excellent studies notwithstanding, much is unknown concerning the rate and sequence of appearance of individual species of fishes following extirpation in streams. Without an understanding of the processes involved in repopulation, the significance of a fish kill is difficult to ascertain. The purpose of this study was to determine the rate of reestablishment of a fish community after a fish kill caused by a pesticide on 17 May 1971 in Mud Creek, Washington County, Arkansas, and to define the ecological factors influencing the rate of succession. The effects of the chemical were short-lived, and all of the damage was done as the substance moved downstream. Mud Creek, a small upland tributary of Clear Creek in the Illinois River drainage, is typified by low flows and nearly isolated pools in late summer and fall and high flows in the spring. It has a lower gradient and more muddy bottom than most small Ozark upland creeks. Mud Creek is normally 1-8 m wide and some pools have a depth of I•A m. MATERIALS AND METHODS 80 TRANS. AMER. FISH. SOC., 1974, NO. 1 Clec•r C "ee• I N 1 Km F1GUaE 1.--Map of Mud Creek showing its relationship to Scull Creek and Clear Creek. The distance A-D represents the kill zone, and B-C depicts the sampling area. the kill on 18 May (Fig. 2). The dead fishes in general had dark gills, distended bellies, expanded opercula, open months, and their heads were turned upward. Most of the moribund fishes were belly side up with their opercula and fins moving slowly. At the confluence of Mud Creek and Clear Creek, stonerollers (Campostoma aaomalum) and orangethroat darters (Etheostoma spectabile), two typically bottom dwelling fishes, were observed swimming in circles at the surface with no apparent equilibrium, and some were running up onto rocks out of the water. Repopulation of the sample area ...