The diagnostic value of sentinel lymph node mapping (SLNM) in patients with colorectal cancer (CRC) is controversial. Prognostic factors for CRC must be detected to improve its treatment. A PubMed query (key words: colorectal cancer, sentinel node) provided 182 studies on the sentinel lymph node (SLN) for CRC, the abstracts of which were reviewed. Altogether, 48 studies dealing with the diagnostic value of SLNM were selected from PubMed, and 6 other studies were retrieved from reviews. We compared the diagnostic value of SLNM with that of conventional histopathologic examination. We used the diagnostic accuracy odds ratio (DAOR) method. Because of significant heterogeneity, we chose the random effect model (Der Simonian and Laird). Statistics were performed on 33 studies, including 1794 patients (1201 colon and 332 rectum cancers). The mean SLNM failure rate was 10%. The global sensitivity and specificity of the SLNM were, respectively, 70% and 81%. The pooled DAOR was 10.7 (95% confidence interval 7.0-16.5). That means that a patient whose SLN is invaded has 10.7 times more risk to be node-positive than an SLN-negative patient. Lymphatic mapping appears to be readily applicable to CRC. One of the main reasons for the heterogeneity is the performance of the SLNM by Saha et al., whose data had better sensitivity (90%) than those in other studies. The SLNM technique should be better standardized in future studies. Understanding the cause of false-negative SLNs (9%) is a major issue to resolve before routinely using this technique in CRC management. The prognostic implication of micrometastases found in SLNs requires further evaluation.