Objective To compare the efficacy of 4 frailty screening tools in elderly cancer patients, and provide a basis for selection of appropriate frailty screening tools in clinical settings. Methods A total of 308 hospitalized elderly cancer patients were selected by purposive sampling method, and were assessed for frailty by using the Geriatric 8, the Vulnerable Elders Survey-13, the Abbreviated Comprehensive Geriatric Assessment, the Frailty Phenotype and the Comprehensive Geriatric Assessment, respectively.With the results of Comprehensive Geriatric Assessment being used as the gold standard for frailty diagnosis, the results of 4 screening tools were compared against the gold standard by running Bayes discriminant analysis and Receiver Operating Characteristic （ROC） analysis, in a bid to determine the frailty screening efficacy of the tools in elderly cancer patients. Results The prevalence of frailty in elderly cancer patients ranged from 46.10% to 70.45%.The consistency of frailty screening results yielded from the Geriatric 8,Vulnerable Elders Survey-13,Abbreviated Comprehensive Geriatric Assessment, and Frailty Phenotype, respectively against the frailty results of Comprehensive Geriatric Assessment generated a Kappa value of 0.599,0.813,0.711 and 0.928,respectively （all P＜0.05）.Cross validation accuracy for predicting frailty risk by the 4 tools was 80.5%,90.6%,85.7% and 96.4%,respectively. The area under the ROC curve was 0.867,0.943,0.853 and 0.976,respectively. Conclusion The Frailty Phenotype has the highest frailty screening efficacy, followed by the Vulnerable Elders Survey-13, the Abbreviated Comprehensive Geriatric Assessment, and the Geriatric 8 in descending order.Courtesy of excellent screening efficacy and easiness to use in clinical settings, the Vulnerable Elders Survey-13 is recommended for frailty screening in elderly cancer patients.