PURPOSE: Total mesorectal excision contains two different procedures: autonomi c nerve preservation, and autonomic nerve sacrifice. It is unclear whether auton omic nerve preservation is suitable curative procedure. We clarify the significa nce of autonomic nerve preservation for an advanced lower rectal cancer. METHODS : All 403 patients curatively resected between 1975 and 1999 were clinicopatholo gically studied. Between 1975 and 1984, all patients routinely received total me sorectal excision without autonomic nerve preservation (TME-P(-) group). Since 1985, total mesorectal excision with autonomic nerve preservation has been perf ormed in 81 percent of patients (TME-P(+) group). The remaining patients recei ved TME-P(-) because of suspicious invasion to autonomic nerve plexus. All cli nical and pathologic data were entered into a computer database. Long-term foll ow-up was used to analyze the oncologic and functional results of TME-P(+) gr oup compared with TME-P(-) group. RESULTS: The follow-up rate was 98.1 percen t. In either Dukes A+B or Dukes C disease, the TME-P(+) group did not increas e local recurrence or decrease ten-year disease-free survival compared with the TME-P(-)-group of Period 1975 to 1984. The TME-P(-) gro up of Period 1985 to 1999 had the highest distant metastasis and the lowest surv ival rates than any other groups. Urinary or sexual function was well preserved in the TME-P(+) group. CONCLUSIONS: Autonomic nerve preservation is oncologica lly and functionally excellent and suitable for almost all patients with advance d lower rectal cancer. Intensive chemotherapy is needed for patients whose auton omic nerves were killed in suspicion of nerve invasion. PURPOSE: Total mesorectal excision contains two different procedures: autonomi c nerve preservation, and autonomic nerve sacrifice. It is unclear whether auton omic nerve preservation is suitable curative procedure. We clarify the significa nce of autonomic nerve preservation for an advanced lower rectal cancer. METHODS All 403 patients curatively resected between 1975 and 1999 were clinicopatholo gically studied. Between 1975 and 1984, all patients routinely received total meorectal excision without autonomic nerve preservation (TME-P (-) group). Since 1985, total mesorectal excision with autonomic The remaining patients recei ved TME-P (-) because of suspicious invasion to autonomic nerve plexus. All cli nical and pathologic data were entered into a computer database. Long-term foll ow-up was used to analyze the oncologic and functional results of TME-P (+) gr oup compared with TME-P (-) group. : The follow-up rate was 98.1 percent. In either Dukes A + B or Dukes C disease, the TME-P (+) group did not increas e local recurrence or decrease ten-year disease-free survival compared with the TME- P (-) - group of Period 1975 to 1984. The TME-P (-) gro up of Period 1985 to 1999 had the highest distant metastasis and the lowest surv ival rates than any other groups. Urinary or sexual function was well preserved in the TME-P (+) group. CONCLUSIONS: Autonomic nerve preservation is oncologica lly and functionally excellent and suitable for almost all patients with advance d lower rectal cancer. Intensive chemotherapy is needed for patients whose auton omic nerves were killed in suspicion of nerve invasion .