Summary C3H/He mice bearing SCC VII tumours received 5-bromo-2'-deoxyuridine (BrdU) continuously for 5 days via implanted miniosmotic pumps in order to label all proliferating (P) cells. The tumours were then heated at 400C for 60 min. At various time points after heating, tumour-bearing mice were irradiated while alive or after being killed. Immediately after irradiation, the tumours were excised, minced and trypsinized. The tumour cell suspensions obtained were incubated with cytochalasin-B (a cytokinesis blocker), and the micronucleus (MN) frequency in cells without BrdU labelling, which could be regarded as quiescent (Q) cells, was determined using immunofluorescence staining for BrdU. The MN frequency in the total (P+Q) tumour cell population was determined from the irradiated tumours that were not pretreated with BrdU. The MN frequency of BrdU unlabelled cells was then used to calculate the surviving fraction of the unlabelled cells from the regression line for the relationship between the MN frequency and the surviving fraction of total (P+Q) tumour cells. In general, Q cells contained a greater hypoxic fraction (HF) than the total tumour cell population. Mild heating decreased the HF of Q cells more markedly than in the total cell population, and the minimum values of HFs of both total and Q cell populations were obtained 6 h after heating. Two days after heating, the HF of total tumour cells returned to almost that of unheated tumours. In contrast, the HF of Q cells did not return to the HF level of unheated tumours until 1 week after heating. It was thought that irradiation within 12 h after mild heating might be a potentially promising therapeutic modality for controlling radioresistant Q tumour cells.Keywords: quiescent cell; hypoxic fraction; mild hyperthermia; immunofluorescence staining; micronucleus assay; 5-bromo-2'-deoxyuridine Tumour hypoxia is clearly an important problem, and improved patient responses to radiotherapy can be achieved by treatments that overcome tumour radiation resistance resulting from the presence of hypoxic cells (Overgaard, 1989). It has been firmly established that hypoxic cells impair the radiation responsiveness of almost all animal tumours thus far investigated (Coleman, 1988).The effectiveness of hyperthermia as an adjuvant modality to radiotherapy has been demonstrated (Overgaard et al, 1995). Laboratory experiments using animal tumours showed that heating for 30 to 60 min at relatively high temperatures, i.e. > 43 to 44°C, damages intratumour blood vessels and kills tumour cells (Vaupel, 1990). Additionally, hyperthermia causes direct cellular radiosensitization (Dewey, 1994). However, currently available hyperthermia devices have been ineffective in raising the temperature of human tumours sufficiently to cause such effects. Furthermore, according to clinical results of thermoradiotherapy, correlations between response to hyperthermia and lowest temperatures in tumours have been reported, and the prognostically important temperatures have been < 41C (Valdagni et ...