To evaluate whether surgical resection confers survival advantages in selected patients with resectable locally recurrent rectal cancer, data on 430 patients who underwent R0 resection for primary rectal cancer were prospectively collected over a 14-year period. Resection of recurrent disease was considered R0 when all cancer tissue was resected with microscopically tumor-free surgical margins. Microscopic evidence of disease at resection margins was considered an R1 resection. Recurrent disease was detected in 158 of 430 patients. Local recurrence was found in 91 patients, including (79%) with resection-site relapse only. These patients were considered for surgery unless defined unresectability criteria were met. A total of 35 patients who had abdominoperineal excision following anterior resection were studied retrospectively. Mortality associated with the procedure was 3% and morbidity was 20%. The resection was R0 in 12 patients, while microscopic margins were involved in 23 patients. 10 patients had extended resection of one or several adjacent organs Median operating time and blood loss were 250 min and 500 ml respectively. Median hospital stay was 25 days. 21/23 R1 patients received postoperative radiotherapy. Return to normal activity occurred at 8.2 (SD 4.2) weeks. No patients were lost to follow-up. Overall median survival was 26.4 months; 5-year survival was 25.4%. In spite of several survival predictors at univariate analysis, R0 or R1 resection was the only independent predictor of survival at multivariate analysis (add ratio 112.7, 95% CI 3.6-3500, p=0.007). Median survival rate was not reached at the 146-month follow-up in patients with R0 resection. Median survival rate was 16.6 months in patients with R1 resection. In conclusion, uninvolved microscopic margins produce long-term survivors after surgical resection for locally recurrent rectal cancer.