Abstract This study aimed to test the agreement of the incremental test's physiological responses between tethered running on a nonmotorized treadmill (NMT) to matched relative intensities while running on a conventional motorized treadmill (MT). Using a within-subject crossover design, nine male recreational runners (age = 22 ± 5 years; height = 175 ± 6 cm; weight = 68.0 ± 16.6 kg) underwent two test sessions: one was an incremental intensity protocol on an MT; the other was on an instrumented NMT. Intensity thresholds at $${\dot{\text V}}$$ V ˙ O 2max , respiratory compensation point (iRCP), and lactate threshold (iLT) were registered for analysis, together with $${\dot{\text V}}$$ V ˙ ̇O 2 , $${\dot{\text V}}$$ V ˙ ̇E, ƒ R, and blood lactate concentration (Lac). Comparisons were based on hypothesis testing (Student's T-test), effect sizes (Cohen's d), ICC, and Bland Altman analysis. Statistical significance was accepted at p < 0.05. Attained $${\dot{\text V}}$$ V ˙ O 2max (MT = 52.2 ± 7.3 mL·kg -1 ·min -1 vs NMT = 50.1 ± 8.1 mL·kg -1 ·min -1 ) and $${\dot{\text V}}$$ V ˙ ̇O 2 at iRCP (MT = 46.3 ± 7.2 mL·kg -1 ·min -1 vs NMT = 42.8 ± 9.3 mL·kg -1 ·min -1 ) were not different between ergometers (p = 0.15 and 0.13, respectively), with significant ICCs (0.84 and 0.70, respectively) and Pearson’s correlations (r = 0.87 and 0.76, respectively). The Lac at iLT presented poor agreement between conditions. Significant correlations were found (r between 0.72 and 0.83) for relative power values of i $${\dot{\text V}}$$ V ˙ O 2max (6.56 ± 1.28 W·kg −1 ), iRCP (4.38 ± 1.50 W·kg −1 ), and iLT (4.15 ± 1.29 W·kg −1 ) related to their counterpart obtained on MT. Results show that running on an NMT offers a higher glycolytic demand under the same relative internal load as running on an MT but with a similar aerobic response and correlated intensity determination.