OBJECTIVE Direct measurement of arginine vasopressin (AVP) via immunoassays is not widely conducted, mainly because of technical constraints. Liquid chromatography-tandem mass spectrometry (LC/MS/MS) has been widely used as the gold standard in clinical chemistry. Here, we aimed to develop an MS-based assay to determine human plasma AVP and compare the results with those obtained using a conventional immunoassay. METHODS We developed a protocol using triple quadrupole MS coupled with LC for the measurement of human plasma AVP. Analytical evaluations of the method were performed, and the results obtained using LC/MS/MS and radioimmunoassay (RIA) were compared. RESULTS The lower limit of quantification (LLOQ) for plasma AVP obtained using LC/MS/MS and RIA were 0.2 and 0.4 pg/mL, respectively. Although there was a weak overall correlation between the results obtained using the two different methods, the RIA results did not agree with the LC/MS/MS results, particularly at low concentrations. CONCLUSIONS AVP detection through RIA is not satisfactory compared with that using LC/MS/MS. Diagnostic values of direct AVP measurements must be evaluated based on the results obtained via sensitive and accurate MS-based methods rather than those obtained through RIA.