Predicted maximal oxygen uptake (VO2max) measurements are based on the assumption of linear relationship between heart rate or power output and oxygen consumption during various intensities. To develop more reliable predicted test for soldiers, the purpose of the present study was to compare the results of direct measurements of VO2max to respective predicted values in cycling (military fitness test). The predicted mean (+/- SD) peak oxygen uptake (VO2peak) value was 45.2 +/- 7.7 mL kg(-1) min(-1) during first week, whereas the respective direct value was 44.8 +/- 8.5 mL kg(-1) min(-1). During the ninth week, the predicted and measured mean (+/-SD) VO2max values were 47.4 +/- 6.7 mL kg(-1) min(-1) and 48.7 +/- 7.3 mL kg(-1) min(-1), respectively. The absolute differences between the methods were -0.42 mL kg(-1) min(-1) (p = 0.46) and 1.28 mL kg(-1) min(-1) (p < 0.05), which correspond to relative values of 0.9% and 2.7%, respectively. A Bland-Altman plot of measured VO2max and predicted VO2max showed no significant trend between the mean and the difference of the 2 methods either before (r = 0.14, p = 0.24) or after the basic military training period (r = 0.11, p = 0.36). Intraclass correlation coefficient varied between r = 0.82 to 0.94. In conclusion, the predicted protocol is fairly accurate (+/-3%) and reliable to predict VO2max values in male soldiers but the use for clinical purposes should be considered individually.