Objective To explore the influencing factors of unplanned failure of continuous renal replacement therapy(CRRT) during extracorporeal membrane oxygenation(ECMO), and to build a prediction model and evaluate its efficacy. Methods A total of 45 patients on CRRT during ECMO were selected. The number of unplanned failure cases of CRRT during ECMO was counted, the indexes in the planned and unplanned failure groups were compared, and logistic regression model was used to construct a prediction model, and the area under ROC curve was used to test the prediction effecacy of the model. Results The 45 patients received 343 sessions of CRRT+ECMO combination treatment, with 212 sessions (61.8%) having unplanned failure without definite inducement. Logistic regression analysis showed that CRRT blood flow velocity (OR=0.924), ECMO blood flow velocity (OR=1.706) and ECMO mode (OR=4.764) were predictors of unplanned failure of CRRT during ECMO. The area under the ROC curve of the prediction model was 0.812, the sensitivity is 0.825, the specificity was 0.696, and the maximum Youden index was 0.521, with prediction model fitting odds ratio χ2=10.113(P=0.257). Conclusion The prediction model enjoys good performance. Clinical medical workers are recommended to use the prediction model for monitoring of CRRT+ECMO combination treatment, and take targeted interventions.