ObjectiveTo analyze the value of electromyography in predicting recovery from acute idiopathic facial nerve paralysis.Study DesignRetrospective case-series review.SettingUniversity-based hospital department of otorhinolaryngology/head neck surgery.PatientsThree hundred fifty-five patients with sudden facial paralysis of unknown cause (Bell's palsy).InterventionTreatment consisted uniformly of high-dose prednisolone, dextran, and pentoxifylline. Prognostication was based on electromyography performed not earlier than 10 to 14 days after the onset of palsy. The findings were classified according to Seddon into neurapraxia and axonotmesis/neurotmesis. There is an inherent statement on prognosis in this classification because neurapraxia is presumed to recover completely within 8 to 12 weeks, whereas axonotmesis is most likely to be followed by sequelae.Main Outcome MeasuresFacial nerve function after 6 months.ResultsComplete recovery was predicted correctly in 92.4% of cases. For the relatively rare and therefore principally more difficult predictable event defective recovery prognosis was still accurate in 80.8%.ConclusionThe detection of spontaneous fibrillation in needle electromyography is a reliable sign predicting unfavorable outcome. An accuracy of 80.8% for predicting unfavorable outcome may be sufficient to advise patients what to expect in the course of their facial nerve disorder. However, it seems dubious to build a decision about surgical intervention on such a test, because in the process, unnecessary surgery would be accepted for as much as one fifth of the patient population.