European Journal of Applied Physiology

, Volume 97, Issue 4, pp 424–431

Optimal power-to-mass ratios when predicting flat and hill-climbing time-trial cycling


    • Research Institute of Healthcare Sciences, Simon Jobson, School of Sport, Performing Arts and LeisureUniversity of Wolverhampton
  • S. A. Jobson
    • Research Institute of Healthcare Sciences, Simon Jobson, School of Sport, Performing Arts and LeisureUniversity of Wolverhampton
  • R. C. R. Davison
    • Department of Sport and Exercise ScienceUniversity of Portsmouth
  • A. E. Jeukendrup
    • School of Sport and Exercise SciencesThe University of Birmingham
Original Article

DOI: 10.1007/s00421-006-0189-6

Cite this article as:
Nevill, A.M., Jobson, S.A., Davison, R.C.R. et al. Eur J Appl Physiol (2006) 97: 424. doi:10.1007/s00421-006-0189-6


The purpose of this article was to establish whether previously reported oxygen-to-mass ratios, used to predict flat and hill-climbing cycling performance, extend to similar power-to-mass ratios incorporating other, often quick and convenient measures of power output recorded in the laboratory [maximum aerobic power (W MAP), power output at ventilatory threshold (W VT) and average power output (W AVG) maintained during a 1 h performance test]. A proportional allometric model was used to predict the optimal power-to-mass ratios associated with cycling speeds during flat and hill-climbing cycling. The optimal models predicting flat time-trial cycling speeds were found to be (W MAP m −0.48)0.54, (W VT m −0.48)0.46 and (W AVG m −0.34)0.58 that explained 69.3, 59.1 and 96.3% of the variance in cycling speeds, respectively. Cross-validation results suggest that, in conjunction with body mass, W MAP can provide an accurate and independent prediction of time-trial cycling, explaining 94.6% of the variance in cycling speeds with the standard deviation about the regression line, s=0.686 km h−1. Based on these models, there is evidence to support that previously reported \(\dot{V}\hbox{O}_{2}\)-to-mass ratios associated with flat cycling speed extend to other laboratory-recorded measures of power output (i.e. Wm −0.32). However, the power-function exponents (0.54, 0.46 and 0.58) would appear to conflict with the assumption that the cyclists’ speeds should be proportional to the cube root (0.33) of power demand/expended, a finding that could be explained by other confounding variables such as bicycle geometry, tractional resistance and/or the presence of a tailwind. The models predicting 6 and 12% hill-climbing cycling speeds were found to be proportional to (W MAP m −0.91)0.66, revealing a mass exponent, 0.91, that also supports previous research.


Power supply and demand Cycling speed Maximal aerobic power (W MAP) Power at ventilatory threshold (W VT) Average power output (W AVG)

Copyright information

© Springer-Verlag 2006