Introduction – Critical power (CP) and ‘functional threshold power’ (FTP) are indices of aerobic fitness and strong predictors of endurance performance (Smith et al., 1999). Completion time (Tlim) during a cycling time trial (TT) can be predicted on the basis of the CP (asymptote) and W′ (curvature constant) parameters of the power-duration relationship according to the equation Tlim=(W-W′)/CP, where W indicates total work done during the TT. However, the accuracy of this performance prediction when CP and W′ are estimated by on-bike power meters is unclear. The FTP test is a popular fitness test for cyclists but its agreement with CP is also yet to be determined (Gavin et al. 2012). Therefore, the aims of this study were to test the hypotheses that: 1) 16.1 km TT performance on the road would be accurately predicted from laboratory-based estimates of CP and W′ using on-bike power meters; and 2) FTP would be significantly correlated with CP. Methods – Following ethical approval, 13 competitive male cyclists (VO2 peak of 63.48 ± 6.54 ml·kg-1·min-1) completed 4-5 TT exercise bouts for the estimation of CP and W′, and a 20 min TT (FTP test). These trials were completed in the laboratory with the participant’s individual competition bike loaded onto a static trainer for rear wheel power output measurement (PowerTap, CycleOps). Subjects also performed a 16.1 km TT on the road on one occasion with the same power measuring device. 16.1 km TT performance was predicted using the equation Tlim= (W-W′)/CP. Results – CP and W′ derived from the power-time relationship were estimated at 277 ± 42 W and 19.5 kJ ± 6.5 kJ, respectively, while FTP was 278 ± 42 W. FTP and CP were not significantly different (P>0.05) and these variables were significantly correlated (r=0.96, P<0.05). Predicted 16.1 km TT performance (27.32 ± 3.29 min) was within 4.4% of, and not significant different to, actual TT performance (26.66 ± 2.17 min, P>0.05, see figure 1). Discussion – Our findings suggest that CP and W′, as derived from mobile power meters in the laboratory, may be used to accurately predict 16.1 km cycling TT performance on the road. In addition, we observed a close agreement between FTP and CP, suggesting that the FTP may provide a time efficient, single-test estimate of the CP. These observations may help inform, and improve the accessibility of, testing procedures that can be employed to predict TT performance in competitive cyclists.