A series of our studies on the development of the Dementia Assessment Sheet for Community‐based Integrated Care System 21‐items (DASC‐21) were reviewed. Study 1: to examine the distribution of scores and internal reliability of DASC‐21, trained nurses and researchers visited the homes of 1341 participants aged 65 years and older and living in the community. The nurses interviewed the participants and, when possible, their family members, to complete the DASC‐21. Then, the Mini‐Mental State Examination was carried out to select the participants of study 2. Study 2: to examine the concurrent and discriminant validity, experienced psychiatrists and psychologists, who were blind to the findings of study 1, visited the homes of 131 subjects who were selected from the participants of study 1 and completed the Clinical Dementia Rating (CDR), the Mini‐Mental State Examination and the Frontal Assessment Battery (FAB). Cronbach's coefficient alpha of the DASC‐21 was 0.808–0.950. Scores of the DASC‐21 significantly correlated with CDR total and box scores, Mini‐Mental State Examination and Frontal Assessment Battery. In an analysis of variance, CDR had the main effect on the score of the DASC‐21. Receiver operating characteristic analysis showed that the DASC‐21 had sufficient discriminatory ability between dementia (CDR1+) and non‐dementia (CDR0 or CDR0.5; area under the curve = 0.804–0.895). When using a cut‐off point of 30/31, sensitivity was 83.3–94.1% and specificity was 77.3–86.4%. The DASC‐21 has sufficient reliability and validity as a tool to evaluate impairments in daily functioning and in cognitive functions, to detect dementia, and to assess the severity of dementia in the community. Geriatr Gerontol Int 2016; 16 (Suppl. 1): 123–131.