Use of maximal aerobic power ((V) over barO(2max)) testing, which requires subjects to exercise to physiological limits, may deter eligible candidates from volunteering for trials and may also be contraindicated in patients suffering from various medical illnesses. An alternative to maximal testing is submaximal testing. The Aerobic Power Index, which represents the aerobic component of the Tri-level Fitness Profile, is a submaximal test that has been shown to be reliable in trained athletes. The purpose of this study was to establish reliability of the Aerobic Power Index, as well as associated variables of (V) over barO(2) (ml(.)kg(-1.)min(-1)) and rate of perceived exertion (RPE), in a group of sedentary subjects. Results for the 20 subjects who participated in a test-retest trial indicated high reliability (ICC r= 0.98; %TEM 3.87 W(.)kg(-1); SEM 0.04 W(.)kg(-1)) for the main measurement outcome of Watts per kilogram (W(.)kg(-1)). Oxygen uptake (ml(.)kg(-1).min(-1)) also demonstrated high reliability (ICC r= 0.92; %TEM 4.63 ml(.)kg(-1.)min(-1); SEM 0.58 ml(.)kg(-1.)min(-1)), as did RPE (ICC r= 0.97, %TEM 7.78; SEM. 0.63). Limits of agreement were -0.02+/-0.16 W(.)kg(-1), -0.41+/- 2.31 ml(.)kg(-1.)min(-1) for (V) over barO(2) and -0.05+/-2.10 for RPE. These results indicate that the Aerobic Power Index is a reliable submaximal exercise test for use in sedentary subjects.