Abstract.-Movement patterns of benthic fishes have not been widely studied, but an understanding of those movement patterns may be necessary in assessing the full effects of habitat fragmentation and the need to restore connectivity in fragmented watersheds. We evaluated nonspawning movement of 3,353 marked Potomac sculpin Cottus girardi over a 1-year period on a 3.14-km segment of a third-order stream. In addition, we conducted a fish removal experiment at the pool-riffle sequence scale (90 m) to evaluate the effects of fish density on movement and recolonization rates. The majority of recaptured Potomac sculpin stayed within 30 m of their original capture location throughout the study (81% at 30 d; 64% at 75 d; 53% at 363 d). Upstream (US) movement was greater than downstream (DS) movement at 30 d (US 17%; DS 3%), 75 d (US 30%, DS 6%) and 363 d (US 32%, DS 15%). After 75 d, 10% of the fish that moved upstream moved more than 250 m (21% after 363 d). The maximum distance moved was 1,711 m in 75 d. There were no significant differences in mean total length between recaptured Potomac sculpin that either moved or did not move or between upstream and downstream movers. The presence or absence of fishes in adjacent 30-m habitat sections influenced upstream, but not downstream, movement of Potomac sculpin. After 30 d, Potomac sculpin population densities and size structure were similar between fish removal and reference sections. We found our results to be similar to those of other studies, which found that, on average, freshwater sculpin are sedentary. However, we identified great dispersal capabilities (.250 m) for a small percentage of Potomac sculpin. The long dispersal distances suggest that this typically sedentary species may need greater habitat connectivity than previously assumed.Information on movement is necessary for assessing the full effects of fragmentation and a species' population needs when restoring connectivity in fragmented watersheds. The short-term and long-term movement patterns of many benthic fishes are unknown, but, in general, home ranges appear to be small. While there has been considerable research on fish movements in general (see reviews by Gowan et al. 1994;Matthews 1998;Rodriguez 2002), most work has focused on salmonids and other sport fish.Previous research has indicated that most freshwater sculpins (Cottidae) do not move frequently or far (see Table 1 for a detailed summary), but no current information has been published on the movement patterns of Potomac sculpin Cottus girardi. Matheson (1979) and Jenkins and Burkhead (1993) do, however, summarize the general ecology and taxonomy of the Potomac sculpin.Bailey (1952) concluded that mottled sculpin Cottus bairdii were relatively sedentary, with typical yearly movements of 50 m. McCleave (1964) estimated the home range of mottled sculpin to be less than 50 m. In two Montana streams, mottled sculpin movement during nonspawning times (June and July) was found to be short (mean ¼ 1.2 m) and haphazard (no preference for upstream or ...