&NA;Records of four hundred thirty‐seven patients with lower and middle rectal cancer who underwent resection for cure at National Cancer Center Hospital from 1969 to 1983 were reviewed. There were significantly lower recurrence rates in the extended excision group compared with the conventional excision group. The recurrence rates between these two groups with Dukes' A were 0 percent (0 of 23)vs.5.2 percent (5 of 96), those with Dukes' B were 6.3 percent (5 of 80)vs.21.9 percent (14 of 64), Dukes' C were 23.6 percent (20 of 89)vs.32.8 percent (28 of 85). The differences between the two groups with Dukes' B and C were statistically significant (P<.05). The cumulative five‐year survival rates in the extended excision group were 94 percent with Dukes' A stage, 88 percent with Dukes' B stage, and 61 percent with Dukes' C stage, compared with 91 percent (Dukes' A), 74 percent (Dukes' B), and 43 percent (Dukes' C) in the conventional excision group. There were also statistically significant differences between the two groups with Dukes' B and C stages (P<.05). Although wide iliopelvic lymphadenectomy was successful as far as decreasing the incidence of local recurrence and also in prolonging survival, there were increased incidences of urine‐voiding failure (loss of sense of bladder being full of urine detected in 39.4 percent of the extended excision groupvs.8.8 percent of the conventional excision group) and sexual impotency (76 percentvs.37.5 percent).