Many researchers have used cycling exercise to evaluate muscle metabolism. Inherent in such studies is an assumption that changes in whole-body respiration are due solely to respiration at the working muscle. Some researchers, however, have speculated that the metabolic cost of torso stabilization may contribute to the metabolic cost of cycling. Therefore, our primary purpose was to determine whether a torso stabilization device would reduce the metabolic cost of producing cycling power. Our secondary purpose was to determine the validity of the ergometer used in this study. Nine male cyclists cycled on a Velotron cycle ergometer at mechanical power outputs intended to elicit 50, 75, and 100% of their ventilatory threshold at 40, 60, and 80 rpm, with and without torso stabilization. Power was controlled by the Velotron in iso-power mode and measured with an SRM powermeter. We determined metabolic cost by indirect calorimetery and recorded power output. Torso stabilization significantly reduced metabolic cost of producing submaximal power (1%), and reduction tended to be greatest at the lower pedaling rates where pedaling force was greatest (1.6% at 40 rpm, 1.2% at 60 rpm, 0.2% at 80 rpm). Power, measured with the SRM powermeter, was strongly correlated with that specified to the Velotron ergometer control unit (R(2) > 0.99). We conclude that muscular contractions associated with torso stabilization elicit significant metabolic costs, which tend to be greatest at low pedaling rates. Researchers who intend to make precise inferences regarding metabolism in the working muscles of the legs may wish to provide torso stabilization as a means of reducing variability, particularly when comparing metabolic data across a wide range of pedaling rates.