Abstract Background Due to the aging of the population the demand for long-term care services is expected to rise during the coming years. For a better planning of health care resources policy makers have to be aware of risk factors associated to nursing home entry (NHE). The present study aims to identify predictors of NHE in a representative sample of Belgian community dwelling older residents. Methods Date from the participants of the Belgian health interview survey (BHIS) 2013 aged 65 years and over were individually linked with longitudinal data from the Belgian compulsory health insurance data (BCHI) over a 5-year period (2012-2017). Institutionalized BHIS participants were excluded, resulting in a final database of 1,927 individuals. A multivariate Cox proportional hazards model was fit to estimate the hazard of NHE. The model examined the hazard of NHE over the follow-up period in function of predisposing, enabling and need variables observed at baseline. All analyses were done using SAS 9.3 taking into account the survey design settings. Results Over the follow-up period, 169 out of 1,927 individuals entered in NH (56% males, mean age =74.7±0.25). Significant predictors of NHE were older age (hazard ratio (HR) =2.40, CI = 1.23-4.67 and HR = 6.19, 95% CI = 2.75-13.92, respectively for 75-84 years and 85+ years compared to 65-74 years), living condition (HR = 4.28, 95% CI = 1.01-18.19 for living alone), severity of limitation in ADLs (HR = 2.61, 95% CI = 1.39-4.88 for moderate limitation and HR = 2.40, 95% CI = 1.10-5.26 for severe limitation, compared to those without limitation). Conclusions Apart from age and living condition, the ADLs limitations were the strongest predictors of NHE. Public health action to reduce NHE of older people should first of all focus on preventive action at middle age which will reduce activity limitations at later age. Key messages Risk profiles for NHE are highly dependent individuals. NH should be more specialized with qualified professionals.