Background Facial motor evoked potential (FMEP) amplitude ratio reduction at the end of the surgery has been identified as a good predictor for postoperative facial nerve outcome. We sought to investigate variations in FMEP amplitude and waveform morphology during vestibular schwannoma (VS) resection and to correlate these measures with postoperative facial function immediately after surgery and at the last follow-up. Methods Intraoperative orbicularis oculi and oris muscles FMEP data from 35 patients undergoing surgery for VS resection were collected, then analysed by surgical stage: initial, dural opening, tumour dissection (TuDis), tumour resection (TuRes) and final. Findings Immediately after surgery, postoperative facial function correlated significantly with the FMEP amplitude ratio during TuDis, TuRes and final stages in both the orbicularis oculi (p = 0.003, 0.055 and 0.028, respectively) and oris muscles (p = 0.002, 0.104 and 0.014, respectively). At the last follow-up, however, facial function correlated significantly with the FMEP amplitude ratio only during the TuDis (p = 0.005) and final (p = 0.102) stages for the orbicularis oris muscle. At both time points, postoperative facial paresis correlated significantly with FMEP waveform deterioration in orbicularis oculi during the final stage (immediate, p = 0.023; follow-up, p = 0.116) and in orbicularis oris during the TuDis, TuRes and final stages (immediate, p = 0.071, 0.000 and 0.001, respectively; follow-up, p = 0.015, 0.001 and 0.01, respectively). Conclusions FMEP amplitude ratio and waveform morphology during VS resection seem to represent independent quantitative parameters that can be used to predict postoperative facial function. Event-to-baseline FMEP monitoring is quite useful to dictate when intraoperative changes in surgical strategy are warranted to reduce the chances of facial nerve injury.