Background: the nodal stage of colorectal cancer (CRC) is a major prognostic factor. The total number of nodes examined from CRC surgical specimens has been associated with improved survival possibly because of increased accuracy in staging. However, systematic sampling of lymph nodes located far from the CRC is time and money consuming and few studies have addressed it's value. Material and Methods: this is a retrospective study of lymph node status from 349 CRC where the mesocolic and perirectal fat was divided into two parts: 1/ close to the tumor (5 cm of both sides of the tumor); 2/ distant from the tumor (above 5 cm of both sides of the tumor). Results: Tumors were located as follows: caecum (61), ascending (31), transverse (33), descending (31), sigmoid (103) and rectum (90). The median number of nodes recovered from the fat close to and distant from the tumor was respectively 13 (range 0-66) and 4 (range 0-33). There were 175 pN0, 104 pN1 and 70 pN2 cases. In all pN1 cases, metastatic lymph nodes were located close to CRC, except for 5 cases in the rectum and 1 case in the colon where metastatic lymph nodes were only present in the distant fat. In pN2 cases, the median number of positive nodes close to and distant from the CRC was respectively 5 (range 3-29) and 0 (range 0-9). Conclusion: In the colon, lymph nodes should be recovered only from the pericolic fat close to the tumor. In the rectum, systematic sampling of nearby and distant lymph nodes is mandatory since rare cases of metastasis arise in distant lymph nodes only. This quantitative restriction of lymph node sampling for CRC surgical specimens might induce a cost effect without having deleterious effects on patient treatment and outcome.