A native population of arctic grayling (Thymallus arcticus) exists in the upper Missouri River drainage of Montana. While formerly abundant, agricultural practices, predation, angling pressure, barriers to mobility, and competition from other species appear to be the causes of decreased presence of artic grayling in the Northwest (Skaar 1989). The purpose of this study was to augment the information base of the swimming ability of arctic grayling and to examine the effect of repeated trials using the same fish. Forty hatchery-raised grayling were separated into two cohorts and reared in an artificial stream. The mean fork length of all fish was 290 mm (SD ± 16). Cohort 1 was tested in a swim chamber experiment once per week for three consecutive weeks. Cohort 2 was tested only once in the swim chamber sixteen weeks later. Each fish was tested by placing it in the chamber initially operating at a low water velocity (30.5 cm s-1) and then regularly increasing the velocity until the fish rested on the screen at the downstream end of the chamber. The maximum water velocity against which each fish held its position was observed, and is reported as the sprint speed (Usprint). Overall the mean Usprint observed was 5.64 BL (body length) s-1 (SD ± 0.78) equal to an absolute velocity of 1.66 m s-1 (SD ± 0.19). Cohort 1 had a mean Usprint of 5.40 BL s-1 (1.62 m s-1) and Cohort 2 had a mean Usprint of 6.33 BL s-1 (1.77 m s-1). There were significant differences, and a generally increasing trend, in Usprint between the successive trials using Cohort 1.