Recently we applied pharmacokinetic-pharmacodynamic modeling to study the relation between the time course of prednisolone concentration and lymphocytopenic effects. A threshold concentration parameter (Cth) was incorporated in the pharmacodynamic model. In the present study the influence of the initial endogenous hydrocortisone (HC) level (HCt = O) was investigated, particularly on Cth estimates. Five healthy subjects received a single p.o. dose of 30 mg of prednisolone on two separate occasions. One dose was given at 8:00 A.M. (data set A), the other at 8:00 P.M. (data set B). A second group of five subjects received 30 mg of prednisolone at 8:00 A.M. on 3 consecutive days, measurements being made after the third dose (data set C). Results from a previous experiment in five subjects who received 30 mg of prednisolone at 8:00 A.M. were reanalyzed in the present study (data set D). Blood samples were taken over a period of 12 hr after dosing for determination of corticosteroid levels and OKT4 positive lymphocyte counts. For data sets B and C, HCt = O tended to be lower and OKT4 counts at t = O tended to be higher in comparison with data sets A and D. Pharmacokinetic parameters of prednisolone were similar in the four experiments. The time course of the OKT4 counts after dosing could be fitted irrespective of the initial counts and HCt = O. The results support the postulate that Cth represents prednisolone replacing lost HC and the conclusion that HC and prednisolone are about equipotent with respect to the lymphocytopenic effect. On the basis of our results and literature data, suggestions are made for optimal immunosuppression with prednisolone.