Between 1988 and 1993, 57 superficial and subsurface tumours of various tumour type were treated with a 430-MHz microwave heating device. Mean (range) tumour depth of the 57 tumours was 3.0 (0.5-6.5) cm. Fifty-four tumours were treated with thermoradiotherapy. Total radiation dose ranged from 20 to 70 Gy with a mean of 53 Gy. For the remaining three tumours, thermochemotherapy was performed. Hyperthermia was given once a week, and a total of 207 heat sessions was administered. Our goal of hyperthermia treatment was to elevate all monitored tumour points > 41 degrees C for > 30 min. The mean (range) number of intratumoral thermometry points was 3.7 (2-6). The goal of hyperthermia treatment was achieved in 49% of the sessions. At the time of maximum tumour regression, complete response was noted in 53% of the tumours treated with thermoradiotherapy. Univariate analysis demonstrated that parameters including tumour type (breast cancer versus others), tumour depth, minimum tumour temperature, average tumour temperature, minimum equivalent time at 43 degrees C, and number of heat sessions achieving the treatment goal significantly affected the tumour response of the combined treatment, while total radiation dose and number of heat sessions were not significant factors for tumour response. Multivariate logistic analysis revealed that only tumour depth (< 3 versus > or = 3 cm) was a significant prognostic factor for tumour response (p = 0.029). Tumour type (breast cancer versus others) and a number of heat sessions achieving the treatment goal (0-1 versus 2-5) were found to be of borderline significance in the multivariate analysis (p = 0.075 and 0.097 respectively). The number of heat sessions achieving a minimum tumour temperature of > 41 degrees C for > 30 min seems a practical thermal parameter that influences tumour response. The present study indicates the importance of quality and quantity of heat session on the treatment outcome of thermoradiotherapy.