In patients with radically resected colorectal carcinoma, lymph node involvement is particularly important for a good prognosis and adjuvant therapy. The number of such lymph node recoveries is still controversial, with recommendations ranging from 6 to 17 nodes. The aim of this study is to determine if a specified minimum number of lymph nodes examined per surgical specimen can have any effect on the prognosis of patients who have undergone curative resection for T(2-4)N(0)M(0) colorectal carcinoma. Between September 1999 and January 2005, a total of 366 patients who underwent radical resection for T(2-4)N(0)M(0) colorectal carcinoma were retrospectively analyzed in a single institution. All specimen segments were fixed, with node identification performed by sight and palpation. We excluded 186 patients who received postoperative adjuvant chemotherapy via oral or intravenous transmission to prevent possible chemotherapeutic effects on patients' prognosis; therefore, a total of 180 patients with T(2-4)N(0)M(0) colorectal carcinoma were enrolled into this study. After the pathological examination, a mean of 12 lymph nodes (range 0-66) was harvested per tumor specimen. No postoperative relapse was found in this group, where the number of examined lymph nodes was 18 or more. Univariate analysis identified the size of the tumor, depth of invasion, grade of tumor, and number of examined lymph nodes, which were significantly correlated with postoperative relapse (all P < 0.05). Meanwhile, both the depth of tumor invasion and the number of harvested lymph nodes were independent predictors for postoperative relapse (P < 0.05). The 5-year overall survival rate of T(2-4)N(0)M(0) colorectal carcinoma patients who had 18 or more lymph nodes examined was significantly higher than those who had less than 18 nodes examined (P = 0.015). Nodal harvest in patients undergoing radical resection for colorectal carcinoma was highly significant in the current investigation. Our results suggest that harvesting and examining a minimum of 18 lymph nodes per surgical specimen might be taken into consideration for more reliable staging of lymph node-negative colorectal carcinoma.