In sedation of dental patients with moderate or severe mental retardation, it is difficult to identify the optimum sedation level and to maintain it appropriately. Moreover, many patients have concomitant epilepsy and are medicated with oral antiepileptic drugs (AEDs), which influence the drug-metabolizing enzymes. In particular, valproate (VPA) has been demonstrated to inhibit propofol metabolism in vitro. Therefore, the objective of the present study was to investigate the clinical influence of oral VPA on the required dose of propofol for sedation, with use of a prospective cohort study design. We studied 45 patients with moderate or severe mental retardation who underwent dental treatment under sedation. Propofol was infused, and sedation was maintained at the same level in all patients using a bispectral index (BIS) monitor. After the completion of treatment for the scheduled patients, patients were divided into those with oral VPA treatment (VPA group: 20 patients) and without any oral antiepileptic treatment (control group: 25 patients). The propofol dose required for sedation and times to the recovery of the eyelash reflex and spontaneous eye opening were evaluated. The median required propofol doses in the VPA and control groups were 4.15 (range 1.975.88) and 5.67 (2.927.17) mg/kg/h, respectively. We observed a statistically significant difference between the two patient groups with respect to median VPA dose (p < 0.01). However, no statistically significant differences were noted in the time until eyelash reflex recovery or spontaneous eye opening between the two groups. The results suggest that oral VPA reduces the dose of propofol required for sedation during dental treatment in patients with moderate or severe mental retardation.