No dependable method has yet been established to find time-dependent concentrations of substances injected into the intervertebral disk. This study investigated the feasibility of microdialysis in the measurement of local concentrations of a low-molecular weight drug in the human lumbar disk. A quasi-static experiment and a dynamic computer finite element simulation were used to study the spread of lidocaine in the lumbar disk. Fresh-frozen cadaveric lumbar motion segments were immersed in a 0.1% lidocaine HCl solution for 6 days prior to continuous microdialysis sampling for 30 min at the posterolateral annulus. Samples were collected every 10 min, for a total of three samples per probe. To maintain quasi-static conditions, where the output of lidocaine was equal to the diffusion rate, the microdialysis flow rate was set to . The finite element model treated the disk as poroelastic tissue under compressive load and introduced 1 ml of 4% lidocaine into the nucleus pulposus. Higher microdialysis flow rates suffered from significant losses during consecutive recoveries. Relative recovery in the annulus at was found to be of the initial solution. This was determined to be a result of low diffusivity of lidocaine through tissue. The FEA model predicted low diffusivity of lidocaine and slow transport to the posterolateral annulus if no fissures were present in the annulus. The results from in vitro experiments and computer simulations showed that while microdialysis can take reliable concentration measurements in the posterolateral annulus, probe placement near a fissure is critical if a measurement is to be made immediately following injection of a drug into the nucleus.