Marsh, CE. Evaluation of the ACSM submaximal treadmill running test for predicting V̇o2max. J Strength Cond Res 26(2): 548–554, 2012—The purpose of this study was to assess the validity of the American College of Sports Medicine's (ACSM's) submaximal treadmill running test in predicting V̇o2max. Twenty-one moderately well-trained men aged 18–34 years performed 1 maximal treadmill test to determine maximal oxygen uptake (M V̇o2max) and 2 submaximal treadmill tests using 4 stages of continuous submaximal exercise. Estimated V̇o2max was predicted by extrapolation to age-predicted maximal heart rate (HRmax) and calculated in 2 ways: using data from all submaximal stages between 110 b·min−1 and 85% HRmax (P V̇o2max-All), and using data from the last 2 stages only (P V̇o2max-2). The measured V̇o2max was overestimated by 3% on average for the group but was not significantly different to predicted V̇o2max (1-way analysis of variance [ANOVA] p = 0.695; M V̇o2max = 53.01 ± 5.38; P V̇o2max-All = 54.27 ± 7.16; P V̇o2max-2 = 54.99 ± 7.69 ml·kg−1·min−1), although M V̇o2max was not overestimated in all the participants—it was underestimated in 30% of observations. Pearson's correlation, standard error of estimate (SEE), and total error (E) between measured and predicted V̇o2max were r = 0.646, 4.35, 4.08 ml·kg−1·min−1 (P V̇o2max-All) and r = 0.642, 4.21, 3.98 ml·kg−1·min−1 (P V̇o2max-2) indicating that the accuracy in prediction (error) was very similar whether using P V̇o2max-All or P V̇o2max-2, with up to 70% of the participants predicted scores within 1 SEE (∼4 ml·kg−1·min−1) of M V̇o2max. In conclusion, the ACSM equation provides a reasonably good estimation of V̇o2max with no difference in predictive accuracy between P V̇o2max-2 and P V̇o2max-All, and hence, either approach may be equally useful in tracking an individual's aerobic fitness over time. However, if a precise knowledge of V̇o2max is required, then it is recommended that this be measured directly.