The rat model of treadmill running is an invaluable tool for the investigation of experimentally and pathologically induced alterations in exercise performance. Interpretation of such data often presumes knowledge of the within-rat reproducibility of performance measures; however, the literature is bereft of this information. We tested the hypothesis that within-rat exercise endurance capacity and peak oxygen uptake (VO_2peak) are highly reproducible across 5 measurements spanning ~5 weeks when assessed with treadmill performance protocols. 13 male Sprague-Dawley rats performed 5 graded exercise tolerance tests to fatigue and 5 maximal exercise tests on a motor-driven treadmill for determination of endurance capacity and VO_2peak, respectively. There were no differences (P = 0.47) in average time-to-fatigue among any of the 5 exercise tolerance tests (average range, 45.9 - 52.1 min), and the average within-rat coefficient of variation (CV) over the 5 runs was 0.13. There were no differences (P > 0.05) among the average CVs from any consecutive weekly exercise tolerance tests (range of 4 CVs, 0.06-0.10). As expected with the increase in body mass, relative VO_2peak decreased (average range, 80.1 75.7 ml/min/kg; P < 0.05) throughout the 5 maximal exercise tests. However, there were no differences (P = 0.63) in the average within-rat CVs among any consecutive VO_2peak tests (range of 4 CVs, 0.03-0.04) and the average within-rat CV for all 5 tests was 0.06. The present data demonstrate that, using the protocols described herein, within-rat measurements of endurance capacity and VO_2peak are highly reproducible. These results have significant implications for improving and refining exercise testing and experimental designs.