Objectives: To propose a simple frailty screening tool able to identify frailty profiles. Design: Cross-sectional observational study. Setting: Participants were recruited in 3 different clinical settings: a primary care outpatient clinic (RURAL population, N=591), a geriatric day clinic (DAY-CLINIC population, N=76) and healthy volunteers (URBAN population, N=147). Participants: A total of 817 older adults (>70 years old) living at home were included. Intervention: A 9-item questionnaire (Lorraine Frailty Profiling Screening Scale, LoFProSS), constructed by an experts' working group, was administered to participants by health professionals. Measurements: A Multiple Correspondence Analysis (MCA) followed by a hierarchical clustering of the results of the MCA performed in each population was conducted to identify participant profiles based on their answers to LoFProSS. A response pattern algorithm was resultantly identified in the RURAL (main) population and subsequently applied to the URBAN and DAY-CLINIC populations and, in these populations, the two classification methods were compared. Finally, clinically-relevant profiles were generated and compared for their ability to similarly classify subjects. Results: The response pattern differed between the 3 sub-populations for all 9 items, revealing significant intergroup differences (1.2±1.4 positive responses for URBAN vs. 2.1±1.3 for RURAL vs. 3.1±2.1 for DAY-CLINIC, all p<0.05). Five clusters were highlighted in the main RURAL population: "non-frail", "hospitalizations", "physical problems", "social isolation" and "behavioral", with similar clusters highlighted in the remaining two populations. Identification of the response pattern algorithm in the RURAL population yielded a second classification approach, with 83% of tested participants classified in the same cluster using the 2 different approaches. Three clinically-relevant profiles ("non-frail" profile, "physical frailty and diseases" profile and "cognitive-psychological frailty" profile) were subsequently generated from the 5 clusters. A similar double classification approach as above was applied to these 3 profiles revealing a very high percentage (95.6%) of similar profile classifications using both methods. Conclusion: The present results demonstrate the ability of LoFProSS to highlight 3 frailty-related profiles, in a consistent manner, among different older populations living at home. Such scale could represent an added value as a simple frailty screening tool for accelerated and better-targeted investigations and interventions.