Introduction: The identification of the HR (or RPE) associated with blood lactate concentrations of 2.5 mmol.L-1 (aerobic threshold) (AerT) and 4.0 mmol.L-1 (anaerobic threshold) (AnT) is a common method for defining training intensities. It is often assumed that the HR at AerT and AnT changes' with changes in fitness, much as the power output (Watts:W) associated with AerT and AnT is known to change. Methods: We studied speed skaters (N = 13, 7 male, 6 female) during spring (deconditioned) and fall (conditioned) evaluations, using cycle ergometry (stage duration = 5 min) to determine W, HR, and RPE at AerT, AnT, and at maximal exercise (3000 (female) and 5000 (male) m cycle time trials). Results: In the spring vs fall evaluations, the power output at AerT was 127 +/- 12 vs 162 +/- 9 W (P < 0.05), at AnT was 216 +/- 14 vs 230 +/- 13 W (P < 0.05), and at maximal exercise was 341 +/- 15 vs 364 +/- 19 W (P < 0.05); HR at AerT was 129 +/- 6 vs 130 +/- 7 bpm (P > 0.05), at AnT was 162 +/- 7 vs 164 +/- 7 bpm (P > 0.05), and at maximal exercise was 196 +/- 6 vs 198 +/- 5 bpm (P > 0.05); RPE at AerT was 2.7 +/- 0.9 vs 2.6 +/- 0.8 (P > 0.05), at AnT was 5.3 +/- 1.0 vs 5.3 +/- 0.9 (P > 0.05). Conclusions: These data suggest that although power output at AerT, AnT, and maximal exercise changes significantly with conditioning, there is no systematic change in the associated values for HR and/or RPE used as practical markers of training intensity. Accordingly, a single well-conducted evaluation may allow evaluation of appropriate training markers that may be longitudinally stable.