Our aim was to evaluate the efficacy and tolerability of anticonvulsant agents for the treatment of acute bipolar mania and ascertain if their effects on mania are a "class" effect. We conducted a systematic review of randomized controlled trials (RCTs) with placebo or active comparator, in acute bipolar mania in order to summarize available data on anticonvulsant treatment of mania/mixed episodes. We searched (PubMed/MEDLINE) with the combination of the words "acute mania" and "clinical trials" with each one of the following words: "anticonvulsants/antiepileptics,"valproic/valproate/divalproex,"carbamazepine,"oxcarbazepine,"lamotrigine,"gabapentin,"topiramate,"phenytoin,"zonisamide,"retigabine,"pregabalin,"tiagabine,"levetiracetam,"licarbazepine,"felbamate," and "vigabatrin." Original articles were found until November 1, 2008. Data from 35 randomized clinical trials suggested that not all anticonvulsants are efficacious for the treatment of acute mania. Valproate showed greater efficacy in reducing manic symptoms, with response rates around 50% compared to a placebo effect of 20-30%. It appears to have a more robust antimanic effect than lithium in rapid cycling and mixed episodes. As valproate, the antimanic effects of carbamazepine have been demonstrated. Evidences did not support the efficacy of the gabapentin, topiramate as well as lamotrigine as monotherapy in acute mania and mixed episodes. Oxcarbazepine data are inconclusive and data regarding other anticonvulsants are not available. Anticonvulsants are not a class when treating mania. While valproate and carbamazepine are significantly more effective than placebo, gabapentin, topiramate, and lamotrigine are not. However, some anticonvulsants may be efficacious in treating some psychiatric comorbidities that are commonly associated to bipolar illness.