A new, practical method called microdialysis has been developed for the estimation of free, unbound drug concentrations in the interstitual fluid. This method is based on the use of urea as an endogenous recovery marker. Urea freely distributes throughout the body water compartment, and its plasma concentration can be used as an accurate measure of its interstitial levels. Microdialysis exploits the relationship between the relative recoveries of urea and the recoveries of an analyte that are established during calibration. The calibration is carried out using different probes, varying perfusion rates, membrane surface areas, temperatures, and tissue models. After obtaining a broad range of recoveries, an empirical mathematical model of the relationship between the recovery of the respective analyte and that of urea is formed. During in vivo experiments, a microdialysis probe is placed into a tissue or fluid of interest, and the urea recovery is monitored by comparing the corresponding dialysate and plasma urea levels. From the calculated urea recoveries, correlation equation, and analyte dialysate concentrations, the extracellular concentration of the unbound analyte in the tissue or fluid can be estimated at any given time point. The purpose of this study is to describe the method of microdialysis and demonstrate its applicability in vitro using two model compounds, theophylline and 3'-azido-3'-deoxythymidine (AZT), in the rat plasma.