BACKGROUND We assessed the feasibility of a robot-assisted laparoscopic radical prostatectomy (rLRP) programme through a review of our early experience. METHODS Seventeen patients underwent rLRP between 1 February 2003 and 31 December 2003 at Singapore General Hospital. All patients had histologically confirmed adenocarcinoma on prostate biopsy and a negative bone scan. The Da Vinci robot was employed. The Montsouris technique was used for our first eight patients, and the Vattikuti Institute Prostatectomy technique was used for all subsequent patients. We studied perioperative parameters and early surgical outcome prospectively. RESULTS The mean age at diagnosis was 63.9 +/- 5.6 years. The median Gleason sum was 6 (range, 5-9), and mean pretreatment prostate-specific antigen level was 10.5 +/- 5.4 ng/mL. The mean set-up time was 34 +/- 18 minutes, and mean dissection time was 247 +/- 43 minutes. Perioperative blood loss averaged 494 +/- 330 mL, and three patients required blood transfusion. Normal diet was resumed after 1.7 +/- 0.6 days. The mean duration of bladder catheterization was 9.8 +/- 6.1 days, and mean hospital stay was 2.7 +/- 1.3 days. There was no perioperative mortality or major complications, and no conversion to open radical prostatectomy. From Case 9 onwards, there was significant reduction in operating time (284 vs 215 minutes), blood loss (650 vs 400 mL) and hospital stay (3.8 vs 1.8 days). CONCLUSIONS rLRP is feasible in a practice with a low volume of radical prostatectomies. Significant improvement in perioperative parameters occurs after the first eight cases. This technique confers the benefits of enhanced precision and dexterity for complex laparoscopic work in the pelvic cavity.