A deep understanding of the dietary patterns and nutrient intake is important for assessment of possilbe nutritional risk and for establishing nutrition improvement strategies. This study was conducted toexamine the dietary characteristics of a nutritionally poor elderly group compared to the middle-and highly-nourished group. Elderly participant was recruited from local elderly centers in Suwon city in 1998. Trained dietitians interviewed 119 elderly(35 males, 84 females) aged 60 years and over for collecting dietary data(24-hour recall) and related variables. Male and female subjects were grouped into high, middle, and low according to the mean nutrient adequancy ratio(MAR) tertiles. An analysisof the percentage of RDA(Recommended Daily Allowances of Korea) for each of the 10 nutrients showed that the male low-MAR group consumed below the RDA in all kinds of nutrients, and the female low-MAR group consumed nutrients below the RDA except vitamin C. An evaluation of nutrient density by Index of Nutritonal Quality(INQ) also showed a similar tendency. Thus, the INQ level of the male low-MAR group was significantly lower than the middle-or high-MAR group, especially in protein, vitamin A, thiamin, riboflavin, and phosphorus(p<0.05). Moreover, INQ level of female low-MAR group was significantly lower than that of the high group(p<0.05) in all nutrients. The female low-MAR group's daily food intake were also lower than those of the high-MARgroup in gains, fish, fruits, oil and beverages. The energy distribution from carbohydrates, fats and proteins showed that the male low-MAR group had significantly higher carbohydrate and lower fat proportions compared to each gender high-MAR group, respectively. The male and female low-MAR group had low scores about eating all side dishes. These findings indicate that a moderate increase of the meat/egg/fishes intake was needed by the male low-MAR group for improving nutrition adequacy, and an overall increase of the food quantity and quality was desired for the female low-MAR group. These data could be used for planning a community elderly nutrition program and establishing strategies for tailored guidelines for the individuals.