Frail elderly patients during their hospitalization can benefit of a comprehensive geriatric assessment (CGA) by an inpatient geriatric consultation team (IGCT). This assessment yields recommendations aiming to improve medical and social management during the patient's hospital stay and after discharge. This study examines the socio-demographic profile of patients assessed by the IGCT and describes the type of recommendations, their adherence rate at 3months and their impact in terms of mortality, rehospitalization, and institutionalization. Retrospective, single center study including all patients assessed by the IGCT during the 4first months of 2009. These 151patients fulfilled criteria of either medical and/or social frailty or had been submitted to an inappropriate in-hospital trajectory. Hospital records and telephone follow-up (with informal and/or professional caregivers as well as primary care physicians) were used to monitor the implementation of recommendations up to three months after hospital discharge. Mean age of the 151patients was 85.6years, 63% were women. 94% of patients lived at home, 70% had a non-supportive environment, 85% were frail and 11% had dementia. On admission, 64% of them had an acute functional loss. 93% of patients were admitted via the emergency room where 67% benefited from CGA. The median hospital duration stay was 14days. At discharge, 76% of patients went back home, 12% were institutionalized and 12% had died. Three months after discharge, 18% of the 134patients surviving hospitalization had been readmitted, 26% institutionalized and 12% had died. The CGA yielded an average of 7recommendations (median value). The mean adherence rate was 78%. The majority of medical recommendations concerned ancillary care and standard medical management, showing high adherence rates (95%). Recommendations concerning the management of cognitive problems were less frequent and overall poorly adhered to (62%). Mean adherence rate was maximal (97%) when the number of recommendations was limited to4 and it dropped under 80% with more than 5recommendations. The recommendations concerning management at home were generally followed (60-77%). Regarding to social outcomes, the main recommendation was to mobilize home assistance and its financial support. There was no significant difference between adherence rate in the hospital and in the community. The extent of adherence rate did not correlate with mortality or readmission at 3months.