The aim of this study was to evaluate the validity of a velodrome field test consisting of repeated rides of 2,280 m, with an initial speed of 28 km.h-1 and increments of 1.5 km.h-1 interspersed with 1-min recovery periods until exhaustion. A group of 12 male competitive road cyclists performed maximal cycling tests under velodrome and laboratory conditions. Velodrome oxygen uptake (VO2) and power output were estimated using equations previously published. Physiological responses to the two tests were compared. Relationships between performance in the velodrome and physiological parameters measured in the laboratory were studied. Maximal power output, heart rate and VO2 were similar in the velodrome and the laboratory [372 (SD 50) vs 365 (SD 36) W, 195 (SD 8) vs 196 (SD 9) beats.min-1 and 4.49 (SD 0.56) vs 4.49 (SD 0.46) l.min-1, respectively], while maximal velodrome blood lactate concentration was significantly higher [13.5 (SD 2.1) vs 11.8 (SD 3.1) mmol.l-1]. Velodrome heart rate was higher at submaximal exercise intensities representing 40%, 50% and 60% of maximal aerobic power, and velodrome blood lactate concentration was also higher at 60%, 70% and 80% of maximal aerobic power. The laboratory parameter that showed the highest correlation with the maximal cycling speed in the velodrome was maximal oxygen uptake (VO2max) expressed per unit of body mass (r = 0.93). In addition, the accuracy of different methods of estimation of the metabolic cost of cycling, rolling resistance, air resistance coefficients and VO2max were compared. Significant differences were found. In conclusion, the present results indicated the validity of a velodrome test used to estimate maximal aerobic parameters of competitive road cyclists, as long as the estimation is made using established equations. When road cyclists are tested in the laboratory, physiological values should be expressed per unit of body surface area or body mass, to predict more accurately the cyclist's performance level under specific field conditions.