Abstract: In order to investigate the prominent feature of food and nutrient intake of healthy people over 80 years from the Bama longevous region, dietary survey of healthy people over 80 years from the longevous region and a non- longevous region in Guangxi province was conducted by means of weighing record method coupled with dietary recall method. Partial least squares- discriminant analysis ( PLS- DA) was used to perform pattern recognition so as to analyze the influence of diet on health and longevity from point of view of the coaction of all kinds of nutrients.The results showed that grain, tubers and miscellaneous bean were main food in the longevous region group and food varieties were diversified. The results of pattern recognition showed that there were significant differences in nutrient intake between the longevous region group and control group ( p < 0.05) . Six characteristic nutrients closely related to the elderly people from longevous region were identified by means of PLS- DA model. Among the six characteristic nutrients, the intake of fat, energy, protein and cholesterol were significantly decreased in the longevous region group ( p < 0.05) , and the fold change values were 0.43, 0.75, 0.61 and 0.26, respectively. Another two characteristic nutrients were carbohydrate and vitamin A, whose intake were significantly increased in the longevous region group ( p < 0.05) , and the fold change values were 1.18 and 1.60, respectively.The specific pattern of nutrient intake was supposed to have a positive influence on the formation of longevity phenomenon.