The purpose of this study was to compare ratings of perceived exertion (RPE; Borg's 6-20 scale) at the same exercise intensity, between a standard exercise electrocardiogram (ECG) treadmill stress test (exECG) and two subsequent bouts of treadmill exercise in a cardiac rehabilitation gymnasium. Eleven patients (mean +/- s) 60.8 +/- 6.1 years performed an exECG within 12.1 +/- 7.5 d after myocardial infarction (MI) and commenced their first exercise-based rehabilitation session (gym-1), which included the use of a motorised treadmill, within 5.0 +/- 1.3 d after the exECG. A second gym session (gym-2) was performed within 4.2 +/- 1.3 d of gym-1. Gym-1 and gym-2 treadmill exercise was performed at an intensity that equated to the penultimate testing stage of exECG, and RPE and heart rate were compared at this level between the three sessions of testing. The mean work rate at the penultimate testing stage of the exECG was 6.0 +/- 1.0 metabolic equivalents; approximately 67% of peak work rate. The RPE at this work rate during exECG, gym-1 and gym-2 were 15.8 +/- 2.7, 13.3 +/- 3.4 and 13.0 +/- 3.6, respectively. A repeated measures ANOVA revealed these RPE responses to be significantly different (F (2, 20) = 9.8; p = 0.001). Post-hoc Bonferroni-corrected pairwise t-tests showed significant differences (p < or = 0.008) between exECG and gym-1 and exECG and gym-2 but not between gym-1 and gym-2. There was no significant difference in heart rate between the three testing sessions (p = 0.076) but it showed signs of a similar trend to RPE. The intra-participant agreement in RPE between gym-1 and gym-2 was substantial; intraclass correlation coefficient (ICC(2,1)) = 0.85 (p < 0.001) and in all but one participant, RPE differed by < or =2 scale points. The RPE responses during standardised exECG treadmill testing, in patients soon after MI, are inflated compared to responses at the same treadmill work rate during subsequent cardiac rehabilitation exercise sessions. Caution is advised in using RPE taken from an initial exECG to guide physical activity in MI patients, but introducing RPE at this point contributes to its subsequent reliable use.