Recent evidence from our laboratory suggests that the factors of tumor site, patient gender, microsatellite instability, and mutations are important determinants in the survival benefit associated with adjuvant chemotherapy in Stage III colorectal carcinoma. In the present study we investigated whether these factors, as well as Ki- mutations, were also associated with the extent of nodal involvement in Stage III cancers. Nodal involvement was retrospectively evaluated in a series of 645 patients with Stage III colorectal cancer from Sir Charles Gairdner Hospital. The number of involved nodes was correlated with the clinicopathologic features of gender, age, tumor site, and histologic grade, as well as to the genetic alterations of mutation, Ki- mutation, and microsatellite instability. The median number of nodes examined per tumor was 11 (range, 1-53). Forty-nine percent of cases had one or two involved nodes and 51 percent had three or more involved nodes, the latter feature being associated with significantly reduced patient survival. No differences in the extent of nodal involvement were apparent with respect to tumor site, patient gender, or or Ki- mutation status. Tumors from younger patients (P = 0.025) or with poorly differentiated histology (P = 0.007), were associated with significantly higher nodal burden, whereas the microsatellite instability phenotype was associated with less extensive nodal involvement (P = 0.020). Survival benefits from the use of chemotherapy were apparent for both the low and high nodal involvement groups, although the latter seemed to obtain relatively more benefit. Multivariate analysis of patients treated with chemotherapy found that gender, grade, and microsatellite instability, but not nodal involvement, were independently prognostic for survival. The extent of nodal involvement in Stage III colorectal cancer is related to patient age, tumor grade, and microsatellite instability status, but not to tumor site, patient gender, or Ki- mutation. These results indicate that differences in metastatic nodal burden cannot explain previously observed site, gender, and mutation differences in the response to chemotherapy.