From 1968-1976, 412 patients were operated on for rectal cancers. One hundred fifty-six underwent abdominoperineal resection (APR) and 256 underwent low anterior resection (LAR). One hundred ninety-two underwent en bloc pelvic lymphadenectomy in conjunction with their resection, while 220 patients underwent more conservative or conventional resection. Thirty-day hospital mortality was 1.7%. The cancer-related 5-year survival was 58.8% for all patients. The proportion of patients surviving 5 years after LAR (62.8%) was significantly better than those surviving after APR (52.4%), p = 0.008. Statistically significantly superior survival was observed after extended dissection when compared to conventional resections in Dukes' A, B, and C patients as a whole (63.8 and 54.3%) and in Dukes' C patients in particular. Superiority of en bloc pelvic lymphadenectomy versus conventional resection was observed in all cases of Dukes' Stage C, Astler-Coller Stage C1, Level II (adjacent) lymph nodes, and Level I (proximal) lymph nodes and was most effective in combination with sphincter-preserving operations. Patient groups were compared for bias and/or case selection, using both contingency tables and Cox-based multiple covariant linear regression analysis, and none was found. In the face of current adjuvant therapy, which is of questionable benefit and which carries its own treatment morbidity, en bloc pelvic lymphadenectomy is advocated as an adjunct to the curative operations for rectal cancer. To improve the overall benefit, patients can be selected for pelvic lymphadenectomy as an adjuvant to resection when preoperative examination suggests that the rectal cancer penetrates the bowel wall. Accurate preoperative staging may help to define a more restricted group of patients warranting (pelvic lymphadenectomy) (PLND). A control randomized trial of the effectiveness of PLND is appropriate to further test its value.