<b>Background</b>: We aimed to elucidate the quantitative relationship between the neuromuscular blockade depth and intraoperative motor-evoked potential amplitudes. <b>Methods</b><i>:</i> This prospective, single-arm, open-label, observational study was conducted at a single university hospital in Seoul, Korea, and included 100 adult patients aged ≥19 years undergoing brain tumor removal surgery under general anesthesia. We measured the neuromuscular blockade degree and motor-evoked potential amplitude in the deltoid, abductor pollicis brevis, tibialis anterior, and abductor hallucis muscles until dural opening. <b>Results</b>: The pharmacokinetic-pharmacodynamic model revealed the exposure-response relationship between the rocuronium effect-site concentration and motor-evoked potential amplitudes. The mean motor-evoked potential amplitudes decreased proportionally with increasing neuromuscular blockade depth. As the mean amplitude increased, the coefficient of variation decreased bi-exponentially. The critical ratio of the first evoked response to the train-of-four stimulation (T1)/control response (Tc) thresholds beyond which the coefficient of variation exhibited minimal change were found to be 0.63, 0.65, 0.68, and 0.63 for the deltoid, abductor pollicis brevis, tibialis anterior, and abductor hallucis muscles, respectively. <b>Conclusions</b>: Our results reveal that the motor-evoked potential amplitude exhibits deterioration proportional to the degree of neuromuscular blockade. In light of the observed bi-exponential decline of the coefficient of variation with the motor-evoked potential amplitude, we recommend maintaining a T1/Tc ratio higher than 0.6 for partial neuromuscular blockade.