Mycophenolic acid (MPA) inhibits the enzyme inosine 5'-monophosphate dehydrogenase (IMPDH). Thus, the measurement of IMPDH activity could serve as a specific pharmacodynamic (PD) tool for monitoring MPA therapy. At present, however, monitoring of pharmacokinetic parameters is preferred over that of PD parameters because, in general, PD assays are labor-intensive and poorly reproducible. Currently, cell count or protein concentration is widely accepted as methods to normalize enzyme activity. In the present study, we have attempted to further improve a method for the determination of IMPDH activity to increase the robustness and reproducibility of the IMPDH activity assay itself, without making the assay more labor-intensive. Therefore, several aspects of the IMPDH method were investigated regarding their influence on the reproducibility and also modified to increase the feasibility and consistency of the assay. The isolation of peripheral blood mononuclear cells (PBMCs) of whole blood samples was found to be the most variable step. Normalization on cell count is labor-intensive and at the same time has a poor reproducibility. Determination of the protein content in cell extracts is impaired by contamination with extracellular proteins and non-PBMCs. Alternatively, the intracellular substance adenosine monophosphate (AMP) was investigated to normalize the newly generated xanthosine monophosphate. Among various subject groups, no significant differences in mean AMP concentration were found. To simplify the procedure, PBMCs were diluted to a fixed volume after isolation from sample of whole blood, and the IMPDH activity was normalized to the AMP concentration quantified in the same high-performance liquid chromatography run as xanthosine monophosphate was quantified. The within-run and total imprecision (coefficient of variation) ranged from 4.2% to 10.6% and from 6.6% to 11.9%, respectively. In conclusion, the modified method described here for the measurement of IMPDH activity can be used reliably in multicenter trials and in longitudinal studies to evaluate the additional value of any PD monitoring among a diversity of patients treated with MPA.