Abstract: Genetic variation is the basis for genetic improvement and plant breeding. This study aimed to estimate the genetic variation by cluster analysis among 20 cultivars of wheat, which has been popularized above 6.67hm2 per year in North Anhui last 50 years. Genetic distance based on the agronomic characters of plant height, panicle length, per panicle, panicle setting density, one thousand grains weight, the coat length, grain length and grain width was calculated and clustered using the unweighted pair-group method Arithmetic Averages(UPGMA). The results showed that 20 cultivars could be classified into eight groups. Group I consists of yang wheat12, wan wheat47; GroupⅡconsists of wan wheat33; Group Ⅲ consists of wan wheat48, fu wheat938, ji wheat21, wan wheat44, wan wheat55, li gao 6,yu wheat 49ck2, liu hu 98, yu nong 035, bai nongAK58; GroupⅣconsists of yu wheat49, wan wheat38; GroupⅤ consists of zheng wheat 366, heng guan35; GroupⅥ consists of wan wheat50; GroupⅦ consists of wan meng118; GroupⅧ consists of yu wheat70. The results were consistent with the grouping based on the available pedigree data. Moreever, cultivars with unkonwn pedigree were assigned to the established heterotic groups. The phenotypes can be applied in classifing cultivars. The study initially provides useful information for the selection of parents in the development of wheat cultivars and hybrids as well.