OBJECTIVE Everolimus is a novel macrolide immunosuppressant used in the prevention of acute and chronic rejection of solid organ transplants. Everolimus is being actively investigated worldwide as a non-nephrotoxic alternative for calcineurin inhibitors. Its highly variable pharmacokinetics and narrow therapeutic window make it difficult to maintain an adequate exposure to prevent serious adverse effects. The primary objective of this study was to improve prediction of everolimus systemic exposure in renal transplant patients by describing the pharmacokinetics of everolimus and identifying the influence of demographic factors and a selection of polymorphisms in genes coding for ABCB1, CYP3A5, CYP2C8 and PXR. The secondary objective of this study was to develop a limited sampling strategy to enable prediction of everolimus exposure in an efficient way and to compare it with the widely used trough blood concentration (C(trough)) monitoring. METHODS A total of 783 blood samples were obtained from 53 renal transplant patients who had been switched from a triple therapy of ciclosporin, mycophenolate mofetil and prednisolone to a calcineurin inhibitor-free dual therapy of everolimus (twice daily) and prednisolone. Everolimus blood concentrations were analysed in whole blood using liquid chromatography-tandem mass spectrometry during routine therapeutic drug monitoring targeting an area under the blood concentration-time curve from time zero to 12 hours (AUC(12)) of 120 μg · h/L. A population pharmacokinetic model was developed and demographic factors and genetic polymorphisms in genes coding for ABCB1, CYP3A5, CYP2C8 and PXR were included as covariates. In addition, a limited sampling strategy was developed. RESULTS Maintaining everolimus systemic exposure at an AUC(12) of 120 μg · h/L resulted in low rejection rates but considerable numbers of adverse events and toxicity. Everolimus pharmacokinetics were best described by a two-compartment model with lag-time (oral clearance = 17.9 L/h; volume of distribution of the central compartment after oral administration [V(1)/F] = 148 L and first-order absorption rate constant [k(a)] = 7.36 h-1). Ideal body weight was significantly related to V(1)/F. None of the selected polymorphisms in genes coding for enzymes involved in distribution and metabolism of everolimus had a significant influence on everolimus pharmacokinetics. The pharmacokinetic limited sampling model (C(trough) and whole blood drug concentration at 2 hours postdose [C(2)]) resulted in a significantly improved prediction of everolimus exposure compared with the widely used C(trough) monitoring. CONCLUSIONS A two-compartment pharmacokinetic model with lag-time describing the concentration-time profile of oral everolimus in renal transplant patients has been developed using pharmacokinetic modelling. Ideal body weight significantly influenced V(1)/F of everolimus; however, the selected polymorphisms in genes coding for ABCB1, CYP3A5, CYP2C8 and PXR had no clinically relevant effect on everolimus pharmacokinetics. Everolimus C(trough) and C(2) as a limited sampling model can be used to accurately estimate everolimus systemic exposure, an improvement over the widely used C(trough) monitoring.