Botulinum toxin injected into a muscle may diffuse to nearby muscles thus producing unwanted effects. In patients with hemifacial spasm, we evaluated clinically and neurophysiologically, whether botulinum toxin type A (BoNT-A) diffuses from the injection site (orbicularis oculi) to untreated muscles (orbicularis oris from the affected side and orbicularis oculi and oris from the unaffected side). We studied 38 patients with idiopathic hemifacial spasm. Botulinum toxin was injected into the affected orbicularis oculi muscle alone (at 3 standardized sites) at a clinically effective dose. Patients were studied before (T0) and 3-4 weeks after treatment (T1). We evaluated the clinical effects of botulinum toxin and muscle strength in the affected and unaffected muscles. We also assessed the peak-to-peak amplitude compound muscle action potential (CMAP) recorded from the orbicularis oculi and orbicularis oris muscles on both sides after supramaximal electrical stimulation of the facial nerve at the stylomastoid foramen. In all patients, botulinum toxin treatment reduced muscle spasms in the injected orbicularis oculi muscle and induced no muscle weakness in the other facial muscles. The CMAP amplitude significantly decreased in the injected orbicularis oculi muscle, but remained unchanged in the other facial muscles (orbicularis oris muscle on the affected side and contra-lateral unaffected muscles). In conclusion, in patients with hemifacial spasm, botulinum toxin, at a clinically effective dose, induces no clinical signs of diffusion and does not reduce the CMAP size in the nearby untreated orbicularis oris or contralateral facial muscles.