Radiation doses delivered in high dose rate (HDR) brachytherapy are susceptible to many inaccuracies and errors, including imaging, planning and delivery. Consequently, the dose delivered to the patient may deviate substantially from the treatment plan. We investigated the feasibility of using TLD measurements in the urethra to estimate the discrepancy in treatments for prostate cancer. The dose response of the 1 mm diam, 6 mm long LiF rods that we used for the in vivo measurements was calibrated with the 192Ir HDR source, as well as a 60Co teletherapy unit. A train of 20 rods contained in a sterile plastic tube was inserted into the urethral (Foley) catheter for the duration of a treatment fraction, and the measured doses were compared to the treatment plan. Initial results from a total of seven treatments in four patients show good agreement between theory and experiment. Analysis of any one treatment showed agreement within 11.7%±6.2% for the highest dose encountered in the central prostatic urethra, and within 10.4%±4.4% for the mean dose. Taking the average over all seven treatments shows agreement within 1.7% for the maximum urethral dose, and within 1.5% for the mean urethral dose. Based on these initial findings it seems that planned prostate doses can be accurately reproduced in the clinic. (C) 2000 American Association of Physicists in Medicine.