Objective: To assess the efficacy and safety of levetiracetam (LEV) as adjunctive treatment in children (0–18 years) with focal-onset seizures (FOS) with a larger dataset. Methods: A pooled analysis would be performed for prospective clinical trials and a meta-analysis for controlled studies. Retrospectives studies were also summarized using descriptive statistics. Results: Thirty-one articles (1763 patients) were identified, eighteen prospective self-controlled studies and thirteen retrospective studies. LEV was more effective than placebo, the pooled risk ratios (RRs) and 95% confidence intervals (CIs) for the 50% responder rate, seizure freedom rate and the median percentage reduction rate were 1.98 (1.49–2.63), 5.12 (2.09–12.51) and 3.19 (2.37–4.30), respectively. The overall response rates (ORRs) and 95% CIs were 56% (52%–60%), 14% (9%–19%) and 55% (31%–79%), respectively. For safety assessment, the pooled RRs and 95% CIs for the at least one treatment-emergent adverse events (TEAE) rate and at least one adverse drug reactions related (ADR-related) TEAE rate were 1.03 (0.94–1.13) and 1.45 (1.13–1.86) between two group. The ORRs and 95% CIs were 74% (54%–94%) and 48% (40%–55%). The adverse events significantly associated with LEV were somnolence 2.26 (95% CI 1.30–3.93) and hostility 2.33 (95% CI 1.15–4.70). The most frequency adverse events were pyrexia, headache, nervousness, upper respiratory tract and somnolence. The RRs for withdrawal rate or the ADR-related withdrawal rate were 0.77 (95% CI 0.44–1.38) and 0.91 (0.42–1.98), the ORRs were 17% (5%–28%) and 6% (4%–8%). Conclusion: The meta-analysis suggested that add-on LEV can significantly reduce seizure frequency and fairly tolerated compared to placebo.