The primary objective of this study was to determine whether physiological measurements obtained from one mode of testing and training could be applied to another mode, as in prescribing heart rate (HR) zone training from cyling to running. Secondary objectives were 1) to assess the validity of applying data from incremental testing to steady-state exercise, and 2) to compare breakpoint in respiratory rate (RR) with the conventional method of anaerobic threshold (AT) breakpoint, the ventilatory equivalent for oxygen VE/VO2). Sixteen experienced triathletes performed VO2max testing on a cycle ergometer (CE) and treadmill (TM). In addition, a 30-minute time trial (TT) was performed on a CE. No significant differences were observed between modes of testing for VO2max (CE = 68.4 +/- 11.1 mlxkgxmin, TM = 69.0 +/- 13.2 mlxkgxmin), maximum HR (CE = 177.1 +/- 6.1 bpm, TM = 178.1 +/- 7.4 bpm), or AT (CE AT HR = 153.9 +/- 10.5 bpm, TM AT HR = 157.0 +/- 9.5 bpm). Although the mean difference in AT HR was small (3.1 bpm), a small correlation coefficient (0.321) between the AT for the 2 testing modes resulted in a large total error (TE = 12.1 bpm), indicating limited practical application of training zones between modes of testing. Mean TT HR and mean TT RR were significantly greater than mean AT HR (159.4 +/- 8.9 vs. 153.9 +/- 10.5 bpm) and mean AT RR (37.8 +/- 6.0 vs. 32.4 +/- 3.2 breaths per minute) because of significant "drift" in these 2 variables over time, whereas TT watts and AT watts were not significantly different (249.1 +/- 47.8 vs. 240.6 +/- 71.1 W). Finally, a significant difference and large TE (9.0 bpm) betweenVE/VO2 AT HR and the RR AT HR (153.9 +/- 10.5 and 158.4 +/- 10.0 bpm) may preclude the practical use of the RR breakpoint. From the results of this study, it is recommended that the triathlete perform sport-specific testing to assess training zones for cycling and running. In addition, because both HR and RR "drift" upward with steady-state exercise, AT RR and AT HR determined by incremental testing underestimate steady-state HR and RR. For this reason, monitoring wattage during steady-state exercise may be more appropriate than monitoring HR and RR.