OBJECTIVE: The aim of the current study was to investigate which is the most suitable classification for colorectal cancer, log odds of positive lymph nodes (LODDS) classification or the classifications based on the number of positive lymph nodes (pN) and positive lymph node ratio(LNR) in a Chinese single institutional population. DESIGN: Clinicopathologic and prognostic data of 1297 patients with colorectal cancer were retrospectively studied. The log-rank statistics, Cox's proportional hazards model, the Nagelkerke R(2) index and a Harrell's C statistic were used. RESULTS: Univariate and three-step multivariate analyses identified that LNR was a significant prognostic factor and LNR classification was superior to both the pN and LODDS classifications. Moreover, the results of the Nagelkerke R(2) index (0.130) and a Harrell's C statistic (0.707) of LNR showed that LNR and LODDS classifications were similar and LNR was a little better than the other two classifications. Furthermore, for patients in each LNR classification, prognosis was homologous between those in different pN or LODDS classifications. However, for patients in pN1a, pN1b, LODDS2 and LODDS3 classifications, significant differences in survival were observed among patients in different LNR classifications. CONCLUSIONS: For patients with colorectal cancer, the LNR classification is more suitable than pN and LODDS classifications for prognostic assessment in a Chinese single institutional population.