Objective Intraoperatively, facial nerve (FN) function is assessed with corticobulbar Motor Evoked Potentials (FN-coMEP). MEP-amplitude decrement and increase in motor threshold (MT) serve as warning criteria. A novel threshold criterion for extremity muscle MEP compares final-to-baseline motor threshold levels (BFB-MT) between the operated versus non-operated side. We applied BFB-MT for FN-coMEP with regard to postoperative FN-function. Methods 79 patients (45f; 48 ± 16 yrs., 68% vestibular schwannoma) undergoing cerebellopontine angle tumor surgery were analysed. FN-function was assessed with the House-Brackmann score (HB) pre- and postoperatively at day 1 (d1), 7 (d7), 3 months (3 m) and grouped in mild (HB score increase ⩽ 1) or marked deterioration (HB score increase ⩾ 2). FN-coMEP were elicited with anodal transcranial electric stimulation at C4;C3 referenced to Cz, and recorded from oris and mentalis muscles. A BFB-MT difference ⩾ 20% (operated vs. non-operated side) was considered significant. MT-increase of ⩾ 20 mA and amplitude decrement ⩾ 50% on the operated side served as reference criteria. FN function was correlated to all 3 criteria. Results At d1, 22% of patients (17/79) showed mild HB deterioration, 27% (21/79) marked. At d7, 25% (20/79) showed mild and 16% (13/79) marked deterioration. At 3 m follow-up (available in 68 patients), 94% showed recovery (64/68). Only 4/68 patients showed residual marked FN function deterioration. FN-coMEP were obtained in 95% cases (75/79). Changes according to BFB-MT (35 ± 17 mA) occurred in 17 patients, 16 of these patients suffered a marked deterioration of FN function at d1. BFB-MT correlated significantly with FN function at d1 and d7. The correlation was stronger using the BFB-MT (Spearman correlation: d1: p = 0.726; d7: p = 0.559; 3 m: p = 0.265) compared to absolute threshold increase ⩾ 20 mA (Spearman correlation: d1: p = 0.675; d7: p = 0.528; 3 m: p = 0.275) and amplitude reduction ⩾ 50% (Spearman correlation: d1: p = 0.305; d7: p = 0.378; 3 m: p = 0.118). Conclusion An increase of BFB-MT ⩾ 20% between the operated vs. non-operated side serves as a better predicator for postoperative FN function than absolute mA increase and amplitude reduction. Further studies have to validate this criterion intraoperatively for its real time applicability.