OBJECTIVE To assess the role of laparoscopic lymph node sampling in patients with locally advanced prostate cancer before radical radiotherapy, and to show that the procedure is safe when carried out by urologists with the appropriate training. METHODS Patients were selected prospectively using a combination of prostate-specific antigen (PSA) level, clinical stage and Gleason grade, which predicted a > 20% risk of lymph node metastases. Between October 1998 and March 2001, 50 patients (mean age 65 years, range 54-78) underwent laparoscopic pelvic lymphadenectomy. Age, presenting PSA level, Gleason grade, clinical stage, operative time, length of stay, and any adverse events were recorded. The histology reports were reviewed for the presence of nodal metastases. The selection criteria were compared between the groups with and with no nodal metastases. RESULTS The mean (range) PSA level of the men was 36.8 (4.4-184) ng/mL and the Gleason grade 6.8 (4-9); 58% had stage T3 disease and 12 patients (24%) had lymph node metastases. There were no statistically significant differences in PSA, Gleason grade or clinical stage between those with positive or negative nodes. The mean operative duration was 110 min, although this decreased from 133 min for the first 10 cases; one case was converted to an open procedure. The median (range) postoperative stay was 1 (1-12) days. There were two major (5%) and seven (17%) minor complications. CONCLUSIONS This study confirms that laparoscopic lymph nodes can be sampled safely by urologists with experience in laparoscopic surgery. Patients with a significant risk of metastases can be selected preoperatively using a combination of PSA, clinical stage and Gleason grade. Lymph node sampling in this group is necessary, as there is no other method of reliably identifying the subgroup of patients with metastatic disease in which radical conformal radiotherapy cannot be justified.