Previous studies have shown that valproate is detectable in vitro by 1H magnetic resonance spectroscopy (MRS) at 1.5 T, whereas in patients on valproate monotherapy, no significant dose-dependent valproate signal could be seen. To investigate whether an increased signal-to-noise ratio as provided by higher valproate doses and increased magnetic field strength would enable detection of valproate in vivo, six Wistar rats were examined using volume-selective 1H MRS at 2.34 T. The spectra were analyzed by fitting a linear superposition of the basis spectra of valproate, brain metabolites, and simulated lipid signals. The analysis revealed no significant signal contributions after valproate administration of up to 330 mg/kg body weight. To analyze how underlying mechanisms, such as potential drug interactions with macromolecules, may affect the valproate signal, additional in vitro spectra of valproate were measured before and after adding albumin. The spectra exhibited a strong decrease of the valproate signal with increasing albumin concentration. The results support the hypothesis that in vivo valproate is bound to a high degree to macromolecules and will therefore not be detectable by 1H MRS.