Background and purpose To identify adverse events (AEs) significantly associated with perampanel treatment in double‐blind clinical studies (RCTs). Serious AEs, study withdrawals due to AEs and dose–effect responses of individual AEs were also investigated. Methods All placebo controlled, double‐blind RCTs investigating therapeutic effects of oral perampanel were searched. AEs were assessed for their association with perampanel after exclusion of synonyms, rare AEs and non‐assessable AEs. Risk difference (RD) was used to evaluate the association of any AE (99% confidence intervals) and withdrawals or serious AEs (95% confidence intervals) with perampanel. Results Nine RCTs (five in pharmacoresistant epilepsy and four in Parkinson's disease) were included in our study. Almost 4000 patients had been recruited, 2627 of whom were randomized to perampanel and treated with drug doses of 0.5 mg/day (n = 68), 1 mg/day (n = 65), 2 mg/day (n = 753), 4 mg/day (n = 1017), 8 mg/day (n = 431) or 12 mg/day (n = 293). Serious AEs were not significantly associated with perampanel treatment. The experimental drug was significantly associated with an increased risk of AE‐related study withdrawals at 4 mg/day RD (95% confidence interval) 0.03 (0.00, 0.06) and 12 mg/day RD (95% confidence interval) 0.13 (0.07, 0.18). Of 15 identified AEs, five (dizziness, ataxia, somnolence, irritability and weight increase) were found to be significantly associated with perampanel and one (seizure worsening) was significantly associated with placebo. Conclusions Vestibulocerebellar AEs (dizziness, ataxia), sedative effects (somnolence), irritability and weight increase were significantly associated with perampanel treatment.