Endoscopic laser ablation is a new treatment for benign prostatic hypertrophy. The objectives of this study were to determine the appearances of the prostate on MR images obtained during, 1 week after, and 3 months after this procedure and to determine if a correlation exists between the MR findings and the clinical outcome. Such appearances could then be used to guide the application of laser energy during the procedure in order to optimize the clinical result. Eight consecutive men 58-74 years old with symptoms of bladder outflow obstruction caused by benign prostatic hypertrophy underwent endoscopic laser ablation of the prostate under spinal or epidural anaesthesia. Imaging was done on a 0.5-T Picker Asset system with an endorectal receiver coil and conventional T1-weighted spin-echo, T2-weighted spin-echo, and gradient-recalled-echo sequences. T1-weighted magnetization transfer images were obtained in three patients. Images were obtained preoperatively, after ablation of the left-sided quadrants, immediately after completion of the procedure, and 1 week and 3 months later. Preoperative and 3-month postoperative symptom scores, peak urine flow rates, and bladder residual volumes were studied. Images were visually assessed for signal-intensity changes and the presence of cavitation by three radiologists in conference. The results were quantitatively analyzed by measuring prostatic volumes on the gradient-recalled-echo images and by measuring the width and area of regions of signal-intensity change on the T2-weighted images. MR images made immediately after treatment showed an increase in the volume of the prostate (mean, 34%) and a poorly defined, low-signal-intensity region around the urethra on the T2-weighted images in six patients. This probably represented coagulative necrosis. The prostate was smaller on MR images made 1 week after treatment, and after 3 months the prostate returned to its preoperative size. After 1 week, the low-signal-intensity periurethral region on the T2-weighted images was less obvious, and at 3 months it was replaced in four patients by a well-demarcated low-signal-intensity ring on the T2-weighted and gradient-recalled-echo images. No evidence of cavity formation in the prostate was seen on MR images in any patient. Symptom scores and peak urinary flow rates improved after 3 months, with a significant difference between the mean increase in symptom scores in the patients with and without the periurethral changes seen immediately after treatment. However, we found no significant difference between the mean increase in peak urinary flow rates in the patients with and without periurethral changes seen either immediately or at 3 months after treatment. No statistically significant correlation was found between the amount of prostatic swelling and the improvement in symptom scores or peak urine flow rates. In patients who have had laser prostatectomy, MR imaging shows significant immediate glandular swelling, which may account for the delayed improvement in symptoms reported with this technique. The presence of the periurethral changes immediately after treatment was correlated with a subsequent improvement in symptom scores. After 3 months, no cavity could be seen in the prostate. This may account for the poorer long-term clinical outcome reported with endoscopic laser ablation of the prostate compared with transurethral prostatectomy.