Monitoring of cyclosporine (CsA) in whole blood is complicated by the important intra- and interindividual variabilities in its pharmacokinetics and by its narrow therapeutic range. Consequently, an individual CsA adjusting dosage is necessary to ensure adequate immunosuppression without major toxic effects. This can be achieved by several methods of drug determination, including immunoassays and high-performance liquid chromatography (HPLC). Despite the use of specific monoclonal antibodies, immunoassays tend to overestimate (at least slightly) CsA blood levels due to a certain percentage of cross-reactivity with drug metabolites; so, at the present time, HPLC remains the reference method. The main problems of CsA analysis by liquid chromatographic (LC) methods depends on its physicochemical properties: a low detection wavelength and the need to use the LC column at a high temperature. A sensitive assay for the determination of CsA in whole blood has been developed using C8 solid-phase column extraction followed by normal-phase LC of the sample. This assay allowed determination of 25 ng/ml CsA in whole blood with an acceptable precision (intra- and interassay variabilities ranged from 7.7 to 10.4%, respectively n = 5). Standard curves constructed daily over the concentration range 25 to 400 ng/ml showed good reproducibility [coefficient of variation (CV) = 5.3%; n = 5] and the assay was linear up to 1,000 ng/ml. A chromatographic run was achieved in 5 min and no late eluting peak was observed. This method has proven to be reliable from the French interlaboratory cyclosporine quality assessment scheme and has been used routinely over 2 years to determine residual blood concentrations of liver transplant recipients.