Although the therapeutic effects of heat on tumors have been known for more than a century, only recently has hyperthermia been applied in the treatment of various solid tumors in a scientific fashion. In preliminary clinical trials, heat applied by transurethrally or transrectally placed microwave antennas has been used in the treatment of benign prostatic hyperplasia (BPH) with some success. However, basic information about prostatic temperature distribution and the cellular effects of hyperthermia is lacking. In an attempt to establish the safety of local hyperthermia of the prostate and the precise temperature distribution within the gland, we performed temperature mapping studies in canine prostates. Eight mongrel dogs were anesthetized, and a 16 Fr catheter with three helical coil microwave antennas was placed in the bladder through a perineal urethrostomy. Laparotomy was performed and the bladder opened. The antennas were placed under direct control in the prostatic urethra. Linear array and single point thermometers (Clini-Therm TS1200 thermometry system) were placed (1) within the catheter alongside the antennas, (2) alongside the Foley catheter in the urethra, and (3) longitudinally and radially in the prostate for mapping of tissue temperature. Heating was performed with a 915-MHz ISM frequency Z-80 microprocessor controlled microwave power generator (Prostek 3000, Clini-Therm Corporation, Dallas, TX) using 2-9 watts per channel (6-27 watts total) for 30 min to 1 hr. Baseline body temperatures varied between the individual dogs from 37 degrees to 38 degrees C, but temperature distribution within the prostate was even prior to heating. With relatively low power (6 watts total), temperatures of greater than or equal to 45 degrees C were reached within the catheter. Therapeutic temperatures of approximately 43 degrees C were achieved in the periurethral prostate. Intraurethral temperatures were in general 1-2 degrees C lower than within the catheter. The radial temperature dropoff in the prostate was sharp and about 1 degrees C per 3-mm distance from the catheter. This would limit the area of effectiveness to a periurethral zone of 1 cm in diameter in this model. No unsafe temperature peaks were noted either intraurethrally or between the prostate and the rectum during steady-state conditions. Histologic studies demonstrated an intact rectal wall and varying degrees of prostatic inflammation and/or necrosis despite uniform treatment regimens administered. These studies demonstrate the short-term safety of microwave hyperthermia in the canine prostate. Further studies will be necessary to determine the clinical efficacy and toxicity in men with BPH.