A total of 13 patients with ovarian cancer were studied during the initial two courses of i.v. thio-TEPA treatment they underwent after primary surgery. Following an increase in the dose from 60 to 80 mg for the second course, no sign of saturation of thio-TEPA elimination processes or of formation of the metabolite TEPA occurred, indicating dose-independent pharmacokinetics. Myelosuppression after courses was registered by serial measurements of platelets and leukocytes. The time to platelet nadir was quite uniformly 3 weeks and tended to be longer than that of leukocytes, which averaged 2 weeks but showed greater interindividual variation. Linear regression analyses of pharmacokinetic parameters versus myelosuppression revealed statistically significant correlations between thio-TEPA pharmacokinetics and the percentage of reductions in leukocytes and platelets at their mean nadirs. In contrast, no such correlation could be demonstrated for TEPA despite its greater exposure to the body in terms of AUC. We advocate further investigation of this pharmacokinetic-pharmacodynamic relationship so as to establish individualized dosing of thio-TEPA.