OBJECTIVE: The purpose of this study is to develop an index to measure oral and maxillofacial frailty in older adults as an acceptable screening tool. METHODS: Three hundred adults (aged >/= 50 years) participated in this study by telephone interview. Ten candidate items for the Oral and Maxillofacial Frailty Index (OMFI) and 20 matched detailed items for oral and maxillofacial functions were asked. Information about sociodemographic and oral health-related variables was also obtained. The test-retest reliability of the 10 candidate items was determined. Correlations of the 10 candidate items with the 20 detailed items or sociodemographic and oral health-related variables were analysed to test the validity of the candidate items. To suggest optimal items for the OMFI, five different models with different numbers of items were constructed and evaluated based on their associations with main variables. RESULTS: The overall internal consistency of the 10 candidate items was .704, and the Cronbach's alpha value of each item was .23-.55. The test-retest reliability was excellent. There were significant correlations between the 10 candidate and 20 matched detailed items. The score of the 10 candidate items was significantly different according to gender, education level and oral health-related variables. The final five items for the OMFI were difficulties in chewing, the necessity of water when eating dry food, difficulties in jaw or tongue movements, difficulties in speaking or pronunciation and difficulties in facial expression. CONCLUSIONS: The five items could be used as a screening tool for evaluating oral and maxillofacial frailty in older adults.