Abstract
Review of randomized controlled trials (RCTs) shows that valproate (VPA) is effective against partial seizures with or without becoming secondarily generalized. A number of RCTs show little difference in the efficacy of VPA and carbamazepine in this patient group, particularly where patients are randomized at the time of diagnosis. The length of time that it has taken to arrive at these conclusions emphasizes the importance of large active-controlled RCTs at an early stage in drug development in informing clinical practice.