Background and purpose: Anxiety and depression are common disabling comorbidities in cervical dystonia (CD) and may predispose to social withdrawal and social cognitive impairments. The relationship between social cognition and depressive/anxiety symptoms in CD is under-investigated. Methods: Forty-six CD patients (40 women; mean age AE SD, 55.57 AE 10.84 years) were administered the following social cognition battery: Affect Naming, Prosody Face and Pair Matching subtests from the Wechsler Adult Intelligence Scale IV and Wechsler Memory Scale IV (social perception), reality-known and reality-unknown false belief reasoning tasks (theory of mind), Empathy Quotient and Social Norms Questionnaire 22 (social behaviour), alongside the Benton Facial Recognition Task (non-emotional facial discrimination). Alongside CD severity, the Hospital Anxiety and Depression Scale measured depressive/anxiety comorbid diagnostic status and severity, and the Liebowitz Social Anxiety Scale assessed social phobia. Social cognition tasks were standardized using published normative data and a cutoff of z < À1.5 for impairment. Results: More than 90% of our CD patients performed normally on social perception and social behaviour tests. Performance on impaired belief reasoning (theory of mind) was impaired in 10 of 46 (21.74%); five of 46 (10.87%) were impaired on the Empathy Quotient. Better performance on the Affect Naming task was associated with comorbid anxiety (g 2 = 0.09, medium-tolarge effect size) and greater anxiety, depression and social phobia severity. Worse performance on the Empathy Quotient was associated with comorbid depression (g 2 = 0.11, medium-to-large effect size) and greater depression severity. CD patients had significantly more difficulties with fearful face identification (P < 0.001). Conclusions : Greater social perception abilities in CD patients with more severe anxiety and depression suggest efficient modulation and self-adaptation of social cognitive skills.