Objective To construct a risk prediction models based on RNA-binding proteins(RBP)related long non-coding RNA (IncRNA)，and provide guidance for clinical prediction of survival of patients with colorectal cancer（CRC）.Methods Analysis of CRC data of 398 CRC tissue specimen and 39 normal control samples collected from the TCGA database was conducted, and the list of RBP-related genes were obtained from EBICA（https://ebiac.uk/GOA/）. R software was applied to analyze the differentially expressed genes (DEGs) of RBP, Gene Ontology（GO）function analysis was carried out to perform functional annotation of RBP DEGs, and the IncRNA associated with RBP genes were screened by correlation analysis.355 CRC samples obtained by selection were randomly assigned to train dataset and validation dataset with a ration of 50%:50% using caret package in R. Univariate Coxregression analysis was adopted to screen prognostically significant RBPrelated IncRNA，LASSO Cox regression analysis was used to identify the optimal prognostic RBP-related IncRNA and construct the prognostic model. After the risk scores were calculated using Cox regression coefficient, the relationship between the risk score of low risk group and high risk group and clinicopathologic features as well as the overall survival (OS) was analyzed. Nomogram that combines the independent prognostic factors was constructed to predict CRC patients' 1year, 2years, and 3years OS probability and then was verified in the validation dataset.Results Among RBP DEGs, 327 genes were upregulated while 166 were downregulated. A total of 994 lncRNAs were identified as RBP-related lncRNAs (R2＞0.4, P＜0.001), of which 29 RBPrelated lncRNAs with prognostic value were obtained. Three out of those RBP-related lncRNAs (LINC02474,NKILA,ZEN1-AS1) were significantly associated with the CRC patients’ OS probability and used for the construction of prognostic model. Significant differences in OS probability were obtained between high-risk group and lowrisk group CRC patients. Additionally, a nomogram containing age，gender, stage，TNM system and risk could effectively predict longterm OS probability of CRC patients.Conclusion The 3 identified genes（LINC02474,NKILA,ZEN1-AS1）prognostic models has certain predictive value for the prognosis of CRC and play an essential role in the pathogenesis of CRC.