Background: The number of lymph nodes (LN) examined is a key factor in determining the stage of colorectal cancer. Over the past 20 years, the examination of fewer than 12 LN has been regarded as a poor prognosis. However, surgical procedures and pathologic examination improvements have increased the number of LN examined. This raised the question of reassessing the best cutoff point for staging colon cancer patients undergoing curative surgery. Methods: We reviewed a total of 3,126 eligible patients with stage II-III colon cancer collected from the Yonsei Cancer Center Registry (YCC) database between 2005 and 2015. As a validation cohort, we used clinicodemographic data from 9,604 patients with stage II-III colon cancer identified by the Netherlands Cancer Registry (NCR) between 2006 and 2015. The hazard ratio (HR) obtained from Cox-proportional regression was used to investigate the effect of LNs' yield on 6-year overall survival (OS). The optimal LN yield was evaluated by analyzing every cutoff from 3 to 165. We used the Kaplan-Meier method to compare the effect of higher vs. lower optimal LN yield on a 6-year OS. Univariate and multivariate Cox regressions were conducted. Results: The proportion of patients with fewer than 12 LN examined dramatically decreased between 2005 and 2015 (17.4% in 2005 vs. 0.6% in 2015, P< 0.001). Based on the 12 LN examined as the conventional cutoff number, there was no significant association between 6-year OS and LN yield in all stages II-III patients (HR = 1.21, P= 0.116; n = 3,126), stage II (HR = 1.39, P= 0.068; n = 1,570), and stage III (HR = 1.18, P= 0.297; n = 1,557) colon cancer. Interestingly, a higher cutoff of 20 LN examined was associated with a significant increase in 6-year OS in all patients (HR = 1.51, P< 0.001 in patients with LN examined less than 20). Multivariate Cox-proportional hazard regression showed a significant decrease in 6-year OS in stage II (HR = 1.39, P= 0.026) and stage III (HR = 1.54, P< 0.001) with less than 20 LN yield. Consistent with the result in the YCC database, less than 20 of LN yield was associated with poorer 6-year OS in stage II-III patients (HR = 1.25, P< 0.001; n = 9,604), stage II (HR = 1.43, P< 0.001; n = 4,689), and stage III (HR = 1.13, P= 0.007; n = 4,915) in the NCR, respectively. Conclusions: Less than 20 LN examined is associated with a poorer prognosis in patients with surgically treated colon cancer, compared to the current cutoff value of 12. Further validation of re-evaluating criteria on the inadequacy of LN retrieval in high-risk stage II colon cancer is warranted.