This study is designed to assess the prevalence at risk of malnutrition according to the Mini Nutritional Assessment (MNA) and evaluate the factors influencing on the nutritional risk of the elderly. Three hundred and nine elderly (110 men and 199 women: mean age =74.1) who participated in meal service in the Chung-buk province were investigated. Mean MNA total score was 21.9 and women had significantly lower MNA scores than men (respectively, 21.5 and 22.8). In the mean time mean MNA-SF (Short Form) score was 10.7, respectively 10.6 for the women and 11.0 for the men, with the difference being statistically significant. The MNA classified 33% of the elderly as well-nourished, 61.7% as at risk of malnutrition and 5.3% as overt malnourished. However, MNA-SF categorized the examinees 40.2% as good and 59.8% at nutritional risk. Those who identified as malnourished elderly had significantly lower mean BMI, mid-arm and calf circumference, poorer functional abilities (ADL, IADL) , lower MAR and food habits scores, and higher number of nutrient $leq$ 75% of RDA than those with at risk of malnutrition and well nourished. Also socioeconomic status such as educational level, self-rated economic status, poverty level, and marital status significantly influenced nutritional status. Similar effect was observed in self-rated nutritional status and health status, dental status, appetite change according to MNA score. Stepwise multiple regression analysis indicated that weight loss was the most predictive item in the total MNA and MNA-SF score. It was found that items such as mobility, living status (home vs institution) , mode of feeding, and pressure sores were inappropriate for assessment of the elderly who are able to participate meal service program. Also, some modifications of items in MNA are needed in order to apply to Korean elderly. Even though the MNA seems to be an useful tool to screen those old people at risk of malnourished, a lot of work is still to be done with this assessment tool to secure its reliability.