The best approach to determining the optimal dose of cyclosporine in renal transplant recipients is unclear. In this prospective investigation, CsA pharmacokinetic studies were performed in 45 patients 1,4, and 12 weeks after the initiation of CsA. Data from 104 studies were then combined to analyze the relationship between CsA kinetic parameters and posttransplant clinical events. Random trough levels, used in the day-to-day adjustment of CsA dose, were examined separately in 19 of the 45 study patients. All CsA levels were measured with high-performance liquid chromatography. Results showed: (1) trough CsA levels, obtained by random sampling, or from the kinetic studies, correlated poorly with dose; however, there was a good correlation between CsA dose and maximum concentration (Cmax, r=.39, Pc.001), area under the concentration-time curve (AUC, r=.45, Pc.001), and terminal elimination halflife (r=.43, Pc.OOl); (2) several pharmacokinetic parameters correlated with subsequent rejection episodes; patients with acute rejection within 2 or 4 weeks after study had 15-31% lower Cmax (P<.05) and 13-19% lower AUC (P<.05) compared to those who were rejection- free; and (3) levels of blood constituents known to bind CsA also correlated with rejection, and this correlation was independent of the impact of kinetic parameters on rejection. Altogether, these results suggested that a limited number of CsA pharmacokinetic studies may be more useful than multiple, random trough levels in monitoring CsA therapy.