OBJECTIVES: This paper compares the efficacy and tolerability of perampanel (PER) relative to other recently approved anti-epileptic drug (AEDs) - lacosamide (LCS), retigabine (RTG), and eslicarbazepine (ESL) for the adjunctive treatment of partial onset seizures with or without secondary generalization and specifically in the secondary generalization subgroup. MATERIALS AND METHODS: A systematic literature review of all RCTs of PER and selected AEDs in EMBASE, Medline, and the Cochrane Central from 1998 to January 2011 with an update in PubMed in March 2013 was performed. A network meta-analysis was conducted for 50% responder rate for overall seizures; withdrawal due to adverse events; seizure freedom; and 50% responder rate for secondary generalized seizures. RESULTS: Twelve RCTs (three PER, three LCS, three RTG and three ESL) were included. PER performed significantly better than placebo for 'responder rate' (OR 2.151, 95% CrI 1.348-3.472) and 'seizure freedom' (OR 2.507, 95% CrI 1.067-7.429). When compared to other agents, PER was found to be equally effective. For 'withdrawal due to adverse events', PER had the lowest odds ratio vs. placebo compared with other AEDs. In the analysis for the subgroup of patients with secondary generalization, only four RCTs (three PER and one LCS) met the inclusion criteria for one outcome (responder rate) for LCS. In this subgroup, PER was statistically significantly better than placebo (OR 2.448, 95% CrI 1.088-5.828). CONCLUSION: PER was statistically significantly superior to placebo in responder rate, seizure freedom, and responder rate in the secondary generalization population. Though PER had statistically significant greater withdrawal compared to placebo, it had the lowest ORs vs. placebo, suggesting a superior safety profile among the comparators included in this analysis. In patients with partial onset seizure with secondary generalization, PER had a statistically significant effect on responder rate compared to placebo.