OBJECTIVE: Mitoxantrone (MTO) was administered to patients with advanced breast cancer either as free MTO (f-MTO) or liposomal MTO (1-MTO). The intra- and interindividual variations in serum pharmacokinetics of MTO were analysed. In addition, the excretion of MTO and its metabolite mitoxantrone dicarboxylic acid (MTOD) in urine was determined. METHODS: The concentration of MTO was measured by high-performance liquid chromatography in serum over a period of 24 h and the amount of MTO and the metabolite MTOD excreted in urine over 18 h was determined. Pharmacokinetic parameters of f-MTO and 1-MTO were calculated. RESULTS: 1-MTO had a significantly longer half-life of distribution in the deep (third) compartment and thus a larger area under the curve (AUC) than f-MTO. No difference was found with respect to distribution in the peripheral (second) compartment. The kinetics of MTO in serum did not significantly differ between patients. In four patients repeated pharmacokinetic analyses gave superimposable results. Thus, there was no enzyme induction during therapy. By contrast, two patients with oedema had a much longer mean residence time (MRT) and AUC for MTO in serum. Despite the altered pharmacokinetics of f-MTD and 1-MTO, no toxic adverse effects occurred in these two patients. CONCLUSIONS: f-MTO and 1-MTO exhibited different distribution patterns in the deep compartment with a significantly increased half-life for 1-MTO. There is no need to monitor MTO for treatment of breast cancer patients with f-MTO. In patients with oedema, the MRT of MTO is prolonged. The clinical relevance of this observation is as yet unclear.