Using a recently developed gas chromatography and mass spectrometry method to determine whole-blood cyclophosphamide (CP) and 4-hydroxycyclophosphamide/aldophosphamide (4-HO-CP/AP) concentrations, we investigated their pharmacokinetics in women receiving CP therapy. Patients (n = 18) received one or two courses of CP: (a) a 90-min i.v. infusion (4 g/m2) followed by a 96-h i.v. infusion (6 g/m2) in combination with high-dose thiotepa; or (b) a 96-h i.v. infusion (6 g/m2) in combination with high-dose thiotepa. Whole-blood exposures to CP [area under the whole blood concentration versus time curve (AUCCP)] and 4-HO-CP/AP (AUC4HOCP) between courses 1 and 2 were compared after normalization to dose (g/m2). A nonproportional increase was observed for the AUCCP between the first course [1112 micrometer. h/g/m2 +/- 14% coefficient of variation (CV)] and the second course (1579 micrometer . h/g/m2 +/- 28% CV) (P < 0.001). In contrast, the AUC4HOCP (27 micrometer . h/g/m2 +/- 25% CV) determined for the first course was 29% higher than the AUC4HOCP (21 micrometer . h/g/m2 +/- 26% CV) for the second course (P < 0.01). The interpatient whole-blood exposures to both CP and 4-HO-CP/AP were remarkably consistent in this patient population with percent CVs ranging from 14 to 28%. Because thiotepa (800 mg/m2) was administered simultaneously with CP during the second course of treatment, possible inhibition of CP metabolism by thiotepa was investigated using human liver microsomes in vitro. IC50 values determined for inhibition of CP metabolism in three individual liver donors ranged from 1.0 to 40 micrometer. However, the clinical relevance of this observation has not been established.