LÉGER, LUC A., VACLAV SELIGER, and LUC BRASSARD. Backward extrapolation of Vo2max values from the O2 recovery curve. Med. Sci. Sports Exercise, Vol. 12, No. 1, pp. 24–27, 1980. The purpose of this study was to validate a method for measuring Vo2max in field conditions. In order to do so, Vo2max obtained by backward extrapolation of the O2 recovery curve at time zero of recovery (BE), was compared to the Vo2max measured directly at the end of a continuous multistage test (Exercise). The optimal way to determine Vo2max was by using a single component exponential least-squares regression on the first three or four 20 s recovery values. When a three-way valve was used to collect expired air immediately at the start of recovery (n = 20), BE Vo2max (VVSD = 56.5 ± 7.4 ml kg-1 min-1) was not significantly different from Exercise Vo2max (56.1 V7.6). When subjects (n = 18) connected themselves to another breathing valve immediately at the end of the test to simulate field conditions BE and Exercise Vo2 max were similar (63.9 ± 6.7 and 63.1 ± 7,0, respectively) as long as a 3 s correction was made for the delay caused by this procedure. BE and Exercise Vo2max were 59.8 ± 8 and 59.2 ± 8.1 for the 36 subjects; the Pearson's correlation coefficient was 0.92. BE Vo2max measured on 11 additional runners tested on a track (62.6 ± 13) yielded similar results as Exercise Vo2max measured on the treadmill using the same protocol (61.2 ± 13.3). The backward extrapolation method appears to be valid to measure Vo2 max when the air collecting equipment might otherwise he cumbersome during actual sport and/or laboratory conditions.