The aim of this study was to investigate the effect of constant versus variable power output cycling exercise on subsequent high-intensity, running performance. Eight triathletes completed two testing sessions (in a random order), which required the subjects to perform 30 min of cycling at either, a constant power output (90% of the lactate threshold), or a variable power output with power output alternating every 5 min (+/-20% of the constant workload). Each cycling bout was immediately followed by a high-intensity treadmill run (16.7+/-0.7 km h(-1)) to exhaustion. No significant differences were found for mean metabolic values or power output between cycling conditions. However, a significant (P<0.05) improvement in run time to exhaustion was reported after 30 min of variable cycling (15:09+/-4:43 min) compared to constant cycling (10:51+/-3:32 min). The results of this study demonstrate that, despite similar average physiological responses during 30 min of cycling, variable-intensity cycling results in an improved running performance compared to constant-intensity cycling. It is hypothesised that the reduced power output in the final 5 min of variable cycling protocol may allow recovery before transition, however the mechanisms involved cannot be determined from the current study.