Background: Antipsychotic drugs are associated with increased risk of seizure. Objectives: To evaluate the seizure risk associated with individual antipsychotics and receptor-binding pro fi les of antipsychotics. Methods: Design: A cohort study of the seizure risk in patients with initiating antipsychotic treatment for psychotic or mood disorder. Setting: Using the Taiwan ’ s National Health Insurance Research Database, patients aged ≥ 15 years and used antipsychotic drug between January 1, 2001 and December 31, 2009 were enrolled. Main outcome measures: Firs-ever seizures were identi fi ed through the emergency department visits and hospitalization with a diagnosis of seizure (ICD-9CM code: 333.2 (myoclonus), 345 (epilepsy), or 780.3 (convulsions). Statistic analysis: Cox proportional hazard model and propensity-score matched analyses were used to evaluate seizure risk of individual antipsychotic drug, compared with Risperidone. Comorbid neurological, psychiatric, and medical disorder, concomitant medication use, and health system utilization were adjusted. Results: A total of 1,867 seizures were identified among 298,752 new episodes of antipsychotic treatment. Compared with Risperidone, the risk of seizure was increased for chlorpromazine (Hazard ratio [HR] =1.65, 95% Con fi dence interval [CI] = 1.25-2.17), chlorprothixene (HR = 2.35 (1.27-4.32), clozapine (1.78 [1.22-2.58]), haloperidol (1.67 [1.40-2.00]), prochlorperazine (1.50 [1.19-1.88]), quetiapine (1.22 [ 1.01-1.49]), and thioridazine (2.06, [1.55-2.75]). In addition, we found that antipsychotics with a high binding af fi nity of D2 dopaminergic and alpha 1 adrenergic receptors, and low binding af fi nity of 5HT2C serotoninergic was associated with a increased risk of seizure than other types of antipsychotics.