The high reproducibility of the key physiologic measurements in our study compares favorably with other studies involving patients with heart, lung, musculoskeletal, or renal disease with less severe exercise impairment. For example, in 11 patients with end-stage renal disease and moderate exercise impairment (mean peak V̇o2 of 1,500 mL/min), Koufaki et al14reported CoVs for peak V̇o2, peak HR, and peak O2P of 4.7%, 5.9%, and 6.0%, respectively. In an exacting study of 17 patients with mild heart disease, who were hospitalized, fasted, and tested at the same time each day, Lehmann and Kolling15 reported correlation coefficients of 0.996 for peak V̇o2, 0.928 for peak HR, 0.991 for peak O2P, and 0.995 for V̇e/V̇co2@AT. Their very high correlation coefficients can be partially attributed to the broad range of peak V̇o2 in their patients (803 to 3,927 mL/min). In 16 cardiac patients with moderate impairment (Weber class A for 6 patients, Weber class B for 5 patients, and Weber class C for 5 patients, mean peak V̇o2 of 26.5 mL/min/kg, 17.4 mL/min/kg, and 13.8 mL/min/kg, respectively) tested several times over a period of 3 to 22 months, Janicki et al,13 found CoVs of 5.7%, 4.4%, and 9.2% for peak V̇o2, peak HR, and AT, respectively. In duplicate tests in 11 cardiac patients (4 patients were New York Heart Association class II, and 7 patients were class III), Meyer et al,,19 in patients with mean peak V̇o2 of 13.9 mL/min/kg, found CoVs for peak V̇o2, peak HR, and peak O2P of 4.1%, 1.4%, and 4.4%, respectively.