Objective: To describe the characteristics of “delayed discharge patients” and the factors associated with “delayed discharges”. Material and methods: We performed a 12-month observational study on patients classified as “delayed discharge patients” admitted to an Academic Internal Medicine ward. We assessed the demographic variables, the number and severity of diseases using the Geriatric Index of Comorbidity (GIC), the cognitive, affective and functional status using respectively the Mini Mental Stare Examination (MMSE), the Geriatric Depression Scale (GDS) and the Barthel Index. We assessed the total length of stay (T-LHS), the total inappropriate length of stay (T-ILHS), the median length of stays (M-LHS), the median inappropriate length of stay (M-ILHS) and evaluated the factors associated with delayed discharge. Results: “delayed discharge patients” were 11.9% of all patients. The mean age was 81.9 years, 74.0% were in the IV class of GIC and 33.5% were at the some time totally dependent and affected by severe or not assessable cognitive impairments. The patients had 2584 T-LHS, of which 1058 (40.9%) were T-ILHS. Their M-LHS was 15 days, and the M-ILHS was 5 days. In general the more is the LHS, the more are the ILHS (Spearman’s rho + 0.68, p<0.001). Using a multivariate analysis only the absence of formal aids before hospitalization was independently associated with delayed discharge (F= 4.39, p=0.038). The majority of delays (69%) resulted from the difficulty in finding beds in long-term hospital wards, but the higher M-ILHS (9 days) was evidenced for patients waiting for the Geriatric Evaluation Unit. Conclusions: The profile of patients and the pattern of hospital utilisation suggest to reorient the health care system and develop appropriate resources for the academic functions of education, research and patient care. Keywords Elderly people – Multidimensional assessment - Delayed discharge - Inappropriate hospital stay