BACKGROUND To investigate the prognostic significance of the number of lymph nodes removed in colorectal cancer (CRC) patients with no metastatic lymph node. METHODS The clinicopathological data of 461 CRC patients was analyzed. In order to compare the survival of patients who had fewer lymph nodes removed versus the survival of patients who had 1-3 metastatic lymph node(s), a separate group of 74 N1 disease patients were also included in the study. All patient data were collected prospectively. Kaplan-Meier method was used for calculation and plotting of the survival curves of the patient groups, and log-rank test was used for the comparison of the survival curves. RESULTS Cancer-specific survival (CSS) rates of patients who had 1-7 lymph node(s) and 8-11 lymph nodes removed were significantly worse than those who had 12 or more lymph nodes removed (p = 0.006 and p = 0.037, respectively), while CSS was not significantly different between those who had 1-7 versus 8-11 lymph node(s) removed (p = 0.647); this grouping had independent prognostic significance in Cox analysis (p = 0.006). CSS of patients with N1 disease was not significantly different from those who had 1-7 and 8-11 lymph node(s) removed (p = 0.312 and p = 0.165, respectively), while it was significantly worse than CSS of patients who had 12 or more lymph nodes removed (p = 0.001). CONCLUSIONS In colorectal cancer patients whose removed lymph nodes are non-metastatic, removal of at least 12 lymph nodes will determine the lymph node status reliably.