The relationship between the percentage of heart rate reserve (%HRR) and percentage of oxygen uptake reserve (%VO₂R) has been recommended for prescribing aerobic exercise intensity. However, this relationship was derived from progressive maximal exercise testing data, and the stability of the relationship during prolonged exercise at a constant work rate has not been established. The main aim of this study was to investigate the stability of the %VO₂R-%HRR relationship during prolonged treadmill exercise bouts performed at 3 different constant work rates. Twenty-eight men performed 4 exercise tests: (i) a ramp-incremental maximal exercise test to determine maximal heart rate (HR(max)) and maximal oxygen uptake (VO₂(max)) and (ii) three 40-min exercise bouts at 60%, 70%, and 80% VO₂R. HR and VO₂ significantly increased over time and were influenced by exercise intensity (p < 0.001 and p = 0.004, respectively). A 1:1 relationship between %HRR and %VO₂R, and between %HRR and %VO₂(max), was not observed, with mean differences of 8% (t = 5.2, p < 0.001) and 6% (t = 4.8, p < 0.001), respectively. The VO₂ values predicted from the ACSM running equation were all significantly higher than the observed VO₂ values (p < 0.001 for all comparisons), whereas a difference for HR was observed only for the tenth min of exercise at 80% VO₂R (p = 0.041). In conclusion, the main finding of this study was that the %HRR-%VO₂R relationship determined by linear regression, obtained from progressive maximal exercise testing, did not apply to prolonged treadmill running performed at 3 work rates.