Objective To understand the status and influencing factors of Activity of Daily Living (ADL) ability among the elderly in Wuhan community, and to establish ADL prediction model. Methods Multi-stage cluster sampling method was utilized to select and investigate 822 older adults aged 60 and over from 13 communities in Wuhan. The sociodemographic data and ADL were analyzed statistically. The model was established using binary Logistic regression, the ROC curve of the model was drawn, and the discrimination of the model was evaluated by AUC.Results The ADL damage rate of the community elderly in Wuhan was about 10.58%, and there were significant differences in age, marital status, work, educational background, main complaint memory decline and hobbies between ADL impaired group (n=187) and ADL normal group (n=635) (P<0.05, P<0.01). Logistic regression analysis showed that, older age and memory decline increased the risk of ADL impairment (P<0.05, P<0.01) , while higher educational background and 2 or more hobbies decreased the risk (P< 0.01 for both). The area under ROC curve was 0.742 (P<0.01), indicating the discrimination of prediction model was good, and the Hosmer-Lemeshow test result (P>0.05) indicated the prediction model had a good calibration. Conclusion The ADL impairment of the elderly in Wuhan is in a medium level, and the prediction model has good prediction value. Measures should be taken to prevent ADL decline for those elderly people over 75 years old, with lower educational background and memory decline and less hobbies, so as to promote healthy aging.