BACKGROUND Prednisolone (PLN) is a widely used corticosteroid in a variety of immune-mediated diseases. Treatment regimes generally consist of empirically derived treatment doses, whereas therapeutic response among patients is highly variable. Drug monitoring of serum PLN levels might support a more rational approach to dose selection, yet is invasive and laborious. In analogy to cortisol, salivary PLN may offer a good alternative for serum PLN, being a representative approximation of free serum PLN. The aims of this study were to evaluate the correlation between free serum and salivary PLN levels and to quantify this relationship within a population pharmacokinetic model. METHODS PLN and prednisone (PN) concentrations were measured in 396 samples from 19 healthy volunteers after oral ingestion of 80 mg PLN. Measurements in serum, ultrafiltrate, and saliva were performed with a recently validated liquid chromatography tandem mass spectrometry method. Population pharmacokinetic analysis was performed with nonlinear mixed effect modeling using NONMEM. RESULTS Salivary PLN levels correlated well with free serum PLN levels (r = 0.931, P < 0.01). A weaker correlation was found for PN (r = 0.318, P < 0.01), which may be explained by the finding that salivary PN levels mainly seemed to consist of PLN enzymatically converted to PN. Total and free serum PLN concentrations decreased over time after drug administration and showed a nonlinear mutual relationship, consistent with concentration-dependent protein binding. Modeled PLN pharmacokinetics corresponded with previous reports. Low to moderate interindividual variability was found for V/F and CL/F (coefficients of variation were 13.8% and 14.6%, respectively). Free and salivary PLN showed a nonlinear relationship with total PLN. An equation predicting free serum levels from salivary levels was successfully derived from the data. CONCLUSIONS This study is the first to describe the relationship between salivary and (free) serum PLN using a population pharmacokinetic model. Salivary PLN was found to be a reliable predictor of free and total serum PLN in healthy volunteers. The results of this study encourage further exploration of the use of saliva as a noninvasive and feasible method for drug monitoring of PLN.