Sensitive detection methods and accurate reporting are necessary to determine the prognostic significance of micrometastases (MM) and isolated tumor cells (ITCs) in lymph nodes that drain colorectal cancers (CRCs). This study examined the role of lymphatic mapping (LM) in the application of the new tumor-node-metastasis (TNM) classification for MM and ITC. All patients at the John Wayne Cancer Institute underwent LM immediately before standard resection of primary CRC between 1996 and 2001. Sentinel nodes (SNs) were identified using blue dye and/or radiotracer and were examined by hematoxylin-eosin (H&E) staining, cytokeratin immunohistochemistry, and multilevel sectioning. The comparison group comprised 370 patients whose primary CRCs were resected without LM during the same period at the same institution. LM was successfully performed in 115 of 120 (96%) patients and correctly predicted the tumor status of the nodal basin in 110 of 115 (96%) patients. Thirty-seven patients (32%) were lymph node-positive by H&E, ITC and MM were found in 23 patients (29.4%) whose lymph nodes were negative by H&E. Tumor deposits were found in the SN only in 29 patients (50%). Nodal involvement was identified for 14.3%, 30%, 74.6%, and 83.3% of T1, T2, T3, and T4 tumors, respectively, in the study group, and for 6.8%, 8.5%, 49.3%, and 41.8% of T1, T2, T3, and T4 tumors, respectively, in the comparison group. The study group had a higher percentage of nodal metastases (53% v 36%; P <.01) and a higher incidence of MM and ITC (29.4% v 1.9%; P <.0001). The mean number of lymph nodes found in the study group (14) was also significantly more than the number found in the comparison group (10; P <.00001). Conventional examination of lymph nodes for CRC is inadequate for the detection of MM and ITC as described in the new TNM classification. Thus, LM and focused SN analysis should be considered to fully stage CRC.