The first aim of our study was to analyze the adverse events statistically significantly associated with zonisamide, through a systematic review and meta-analysis of available randomized placebo-controlled trials (RCTs). The second aim was to compare these results with those obtained from an analysis of non-RCTs and observational studies. Randomized controlled trials were identified using Medline (PubMed), EMBASE (Ovid), and Cochrane CENTRAL, from 1990 to September 2012. RevMan version 5.1 and OpenMeta[Analyst] were used for analyses of RCT and non-RCTs, respectively. Six eligible studies with 1184 patients between 12 and 80 years of age were included in RCTs analysis. Fifteen adverse events were investigated. In this first part of the analysis, no adverse events were statistically significantly associated with zonisamide. In the non-RCT analysis, a high incidence of weight loss and headache was found. In RCTs, zonisamide was statistically significantly associated with an increased risk of adverse event-related study withdrawals [RR (99% CI) = 1.81 (1.07–3.08)]. Although our study revealed no statistically significantly associated adverse effects (AEs) with zonisamide, this is very likely a consequence of the small numbers in the RCTs available. The limited data available from the studies appear to reveal no major safety concerns related to zonisamide. However, the high incidence of weight loss and headache in the non-RCTs suggests that these AEs could be of clinical significance.