Reliability of power output during dynamic cycling.

The aims of the present study were to determine the influence of familiarization on the reliability of power output during a dynamic 30-km cycling trial and to determine the test-retest reliability following a 6-week period. Nine trained male cyclists performed five self-paced 30-km cycling trials, which contained three 250-m sprints and three 1-km sprints. The first three of these trials were performed in consecutive weeks (Week 1, Week 2 and Week 3), while the latter two trials were consecutively conducted 6 wk following (Week 9 and Week 10). Subjects were instructed to complete each sprint, as well as the entire trial in the least time possible. Reproducibility in average power output over the entire 30-km trial for Week 2 and 3 alone (coefficient of variation, CV = 2.4 %, intra-class correlation coefficient, ICC = 0.93) was better than for Week 1 and 2 (CV = 5.5 %, ICC = 0.77) and Week 9 and 10 alone (CV = 5.3 %, ICC = 0.57). These results indicate that high reliability during a dynamic 30-km cycling trial may be obtained after a single familiarization trial when subsequent trials are performed within 7 days. However, if cyclists do not perform trials for six weeks, the same level of reliability is not maintained.