Recent morphometric magnetic resonance imaging (MRI) studies suggested the possibility that valproate (VPA) use is associated with parieto-occipital cortical thinning in patients with heterogeneous epilepsy syndromes. In this study, we examined the effect of VPA on the brain volume using a large number of homogenous patients with idiopathic generalized epilepsy. Voxel-based morphometry was used to compare regional gray matter (GM) volume between 112 patients currently taking VPA (VPA+ group), 81 patients not currently taking VPA (VPA- group), and 120 healthy subjects (control group). The VPA+ group showed a significant GM volume reduction in the bilateral cerebellum, hippocampus, insula, caudate nucleus, medial frontal cortex/anterior cingulate cortex, primary motor/premotor cortex, medial occipital cortex, and anteromedial thalamus, as compared to the control group. The VPA- group showed a significant GM volume reduction in the anteromedial thalamus and right hippocampus/temporal cortex, as compared to the control group. Compared to the VPA- group, the VPA+ group had a significant GM volume reduction in the bilateral cerebellum, primary motor/premotor cortex, and medial frontal cortex/anterior cingulate cortex. We have provided evidence that VPA use could result in GM volume reductions in the frontal cortex and cerebellum. Our findings should be acknowledged as a potential confounding factor in morphometric MRI studies that include subjects taking VPA.