To assess the reliability of a 5-min-stage graded exercise test (GXT) and determine the association between physiological attributes and performance over stochastic cycling trials of varying distance.
Twenty-eight well-trained male cyclists performed 2 GXTs and either a 30-km (n = 17) or a 100-km stochastic cycling time trial (n = 9). Stochastic cycling trials included periods of high-intensity efforts for durations of 250 m, 1 km, or 4 km depending on the test being performing.
Maximal physiological attributes were found to be extremely reliable (maximal oxygen uptake [VO2max]: coefficient of variation [CV] 3.0%, intraclass correlation coefficient [ICC] .911; peak power output [PPO]: CV 3.0%, ICC .913), but a greater variability was found in ventilatory thresholds and economy. All physiological variables measured during the GXT, except economy at 200 W, were correlated with 30-km cycling performance. Power output during the 250-m and 1-km efforts of the 30-km trial were correlated with VO2max, PPO, and the power output at the second ventilatory threshold (r = .58–.82). PPO was the only physiological attributed measured during the GXT to be correlated with performance during the 100-km cycling trial (r = .64).
Many physiological variables from a reliable GXT were associated with performance over shorter (30-km) but not longer (100-km) stochastic cycling trials.