To construct a nomogram for overall survival (OS) in colorectal cancer. Using Cox univariate and multivariate analysis, significant variables were plotted into a nomogram. The predictive accuracy and discriminative ability of the nomogram was determined using the concordance index (C-index), receiver operator characteristic curve and calibration curve. CA125 and tumor node metastasis (TNM) were identified as independent predictors for OS in colorectal cancer patients. Area under the curve (AUC) for 1, 3, and 5 years was 0.89, 0.723, and 0.71 respectively in primary cohort, and C-index was 0.702 (95% CI, 0.610 to 0.794) in validation cohort, which indicated a middle degree of accuracy. High-quality calibration plots were seen in the OS nomogram models. Age, CA125, and TNM were identified as poor predictors of colorectal cancer. This nomogram is a valuable predictive tool for the prognosis of colorectal cancer.