Valproic acid (VPA) dosing needs to be individualized for epilepsy patients through therapeutic drug monitoring (TDM). The patients must show up in the clinic at the therapeutic window time to venipuncture sample. Dried plasma spot (DPS) sampling is an alternative way to replace conventional venipuncture sampling. The aim of this study was to develop and validate a DPS-based liquid chromatography-tandem mass spectrometry (LC-MS/MS) method to monitor VPA in a routine clinical laboratory setting. We compare the DPS with the wet plasma method of clinical samples by LC-MS/MS. The method was linear over the dynamic range of 10-200 μg/ml (covering entire therapeutic range) with a correlation coefficient r2 ≥ 0.995. Both the DPS and wet plasma methods were fully validated for the accuracy, precision, recovery, and matrix effect. The analyte stability was examined under conditions mimicking the sample storage, transport, and analysis procedures. A clinical study with epilepsy patients receiving VPA (n = 35) showed that, after correction for hematocrit (HCT), plasma concentrations can be successfully calculated from the DPS quantification results. Passing-Bablok regression coefficients showed no proportional bias between estimated and measured plasma concentrations. Similar agreement was found by Bland-Altman plots. The dried sample could be mailed to the clinical lab to test by regular mail service. So DPS can be used for drug monitoring with self-sampling strategy at the patient's convenient time and place specially for ambulatory patients not attending a clinic.