A great number of investigators have, independently, shown that tumour blood flow is affected by a hyperthermic treatment to a larger extent than normal tissue blood flow. While the majority of the studies on experimental tumours show a decrease and even a lapse in blood flow within the microcirculation during or after hyperthermia, the data on human tumours are less conclusive. Some of the investigators do not find a decrease in circulation, while others do. Obviously, this is an important field of investigation in the clinical application of hyperthermia because a shut down of the circulation would not only facilitate tumour heating (by reducing venous outflow, this reducing the 'heat clearance' from the tumour), but would also facilitate tumour cell destruction. The same holds for alterations that occur subsequently to the circulatory changes, like a heat-induced decrease of tissue pO2 and pH. If the frequently reported circulatory collapse of the tumour circulation could selectively be stimulated by, e.g. acidification or by vasoactive agents, hyperthermic treatment of patients would possibly be greatly facilitated and intensified. In hyperthermic tumour therapy a number of complex processes and interactions takes place, especially when the treatment is performed in combination with radiation therapy. One of them represents the group of processes related to the random probability of cell sterilization of individual tumour cells resulting in exponential survival curves which are typically evaluated with e.g. cell survival assays. This aspect has not been the issue of this paper. The other group of processes deals with the heat-induced changes in the micro-physiology of tumours and normal tissues which, as discussed before, may not only enhance the exponential cell kill, but which may also culminate in vascular collapse with the ensuing necrosis of the tumour tissue in the areas affected. If this takes place, a process of bulk killing of tumour cells results, rather than the random type of cell sterilization. At present it is not clear to what extent the various separate mechanisms contribute to the total effect of tumour control. With all these considerations in mind, one should be aware of the fact that effects, secondary to heat-induced vascular stasis alone will never be efficient enough to eliminate all tumour cells, even though a heat reservoir is created. This is so because some malignant cells will inevitably have already infiltrated normal, surrounding structures and will therefore not be affected by changes in the tumour vascular bed.(ABSTRACT TRUNCATED AT 400 WORDS)