Despite recent evidence suggesting that Bell's palsy is associated with reactivation of alfa-herpes viruses, the disease has been treated empirically, and the use of valacyclovir has not been definitively established. In 2007, two prospective, randomised, placebo-controlled trials evaluating valacyclovir were reported in patients with Bell's palsy. One demonstrated that valacyclovir/prednisolone therapy was statistically more effective than placebo/prednisolone therapy in improving the recovery of patients with Bell's palsy, excluding zoster sine herpete. However, considering the cost-benefit ratio of this treatment and the limitations of virological diagnoses, we recommend that valacyclovir should be used in cases of severe palsy within 3 days after the onset of Bell's palsy.