Mycophenolic acid (MPA) is an immunosuppressive drug partly metabolized to MPA-glucuronide (MPAG), which is pharmacologically inactive. The currently available enzyme-multiplied immunoassay technique (EMIT) has been reported to overestimate MPA plasma concentration in clinical samples when compared with HPLC techniques. The aims of this study were to design and validate a specific liquid chromatography-tandem mass spectrometry (LC-MS/MS) technique for the determination of MPA and MPAG using a low plasma volume and a simple sample preparation procedure; then to compare it with EMIT for the determination of MPA in plasma samples collected over an interdose interval at different posttransplantation periods (days 3, 7, and 30 and after 3 months) in 25 renal transplant recipients orally administered cyclosporine and mycophenolate mofetil twice daily, to investigate the origins of the differences between techniques. The LC-MS/MS technique developed showed limits of quantification (LOQs) of 0.1 mg/L and 1 mg/L for MPA and MPAG, respectively, and was linear, accurate, and precise from these LOQs up to 30 mg/L for MPA and 300 mg/L for MPAG. EMIT gave similar results to LC-MS/MS for spiked quality control samples (in a synthetic matrix or in drug-free plasma) but significantly overestimated MPA levels in clinical samples: EMIT - LC-MS/MS = +61.39% +/- 57.94%, with large variations depending on patients, time elapsed since transplantation, sampling time, and concentration levels. These results confirmed the known overestimation of the EMIT assay compared with a specific method and showed that the magnitude of this overestimation depended on sampling time and time after transplantation.