Direct measurement of human metabolism is recommended for accuracy in individuals’ exercise testing; when direct measurement is not feasible, estimation of metabolism during exercise is acceptable. The American College of Sports Medicine (ACSM) suggests estimating oxygen consumption (VO2) during exercise on a cycle ergometer through a straightforward equation that considers an individual’s body mass and work rate (ACSM, 2009). Despite avoiding complex calculations, this equation does not consider the effect of pedalling rate (PR) on VO2. However, when cycling at a given work rate, PR can be a strong determinant of VO2 (Seabury et al., 1977, Zoladz et al., 2000). We hypothesised that, for sub-maximal exercise on the cycle ergometer, including PR in the ACSM equation would allow more accurate VO2 prediction. Seven healthy male participants (average ± standard deviation: age 26 ± 9 years, height 180 ± 6 cm, weight 76 ± 10 kg) took part in the study, which conformed to the Declaration of Helsinki and had been approved by the local Ethics Committee. Participants’ VO2 was recorded during steady-state exercise on the cycle ergometer at four PR (50, 70, 90, 110 revolutions per minute); each PR was studied at four subsequent sub-maximal external work rates (0, 50, 100, 150 W). Participants rested for a period of at least 40 min between tests. Participants’ VO2 was also predicted as a function of participant’s body mass, work rate and PR (Minetti, 2011): VO2 = 3.5 + 10.8 * WR/BM + 0.0000076 * PR (Equation 1) Here, 3.5 is an assumed constant for resting VO2 (ml O2 kg-1 min-1), WR indicates work rate (W), and BM participant’s body mass (kg); PR is pedalling rate (revolutions per minute), and represents the metabolic equivalent of the internal work. Participants’ VO2 was also predicted through the ACSM equation as a means of comparison. Data were analysed by Student’s paired t-test and linear regression analysis. The recorded VO2 values ranged from 6 to 37 ml O2 kg-1 min-1. Figure 1 shows the VO2 values predicted by Eq. 1 and by the ACSM equation as a function of measured VO2 values. Both sets of predicted VO2 values showed a direct linear relationship with measured VO2 values. When plotted against the measured VO2 values, those predicted by Eq. 1 were closer to the identity line than values predicted by the ACSM equation. We conclude that PR is an important determinant of VO2, and that Eq. 1 improves the accuracy of the ACSM equation for exercise on the cycle ergometer. These preliminary findings need to be confirmed in a larger sample population.