To clarify the phenotypic differences and genetic relationships of the starch-type potatoes， and to provide a research basis and breeding materials for the breeding and quality improvement of new varieties of starch-type potato. This study used 26 starch-type potato germplasm resources as materials. Through field agronomic trait observation， quality trait measurement， and simple sequence repeats （SSR） marker analysis， genetic diversity was analyzed based on both phenotypic and molecular markers. The results showed that the coefficient of variation of the phenotypic traits ranged from 9.15% to 94.14%， with the highest coefficient of variation for the number of branches， and the lowest coefficient of variation for the dry matter content. The genetic diversity index ranged from 1.71 to 2.09， with the highest genetic diversity index for the number of main stems and the lowest for the number of branches. Based on phenotypic traits， the tested materials were divided into four groups： Group I exhibited high starch content， Group II showed greater plant height and main stem number， Group III was characterized by lower dry matter and starch content， and Group IV performed best in terms of stem thickness and branch number. 31 pairs of SSR primers amplified 408 polymorphic loci， with a polymorphic locus rate of 85.18%. On average， each pair of primers amplified 13.16 polymorphic loci， and the polymorphic information content of the primers ranged from 0.1016 to 0.4024. Based on the SSR marker results， the materials were classified into 4 clusters. Cluster I included only Qingshu No. 9， with red tuber skin. Cluster II contained 22 materials， of which 21 were bred in Gansu， Ningxia， Shanxi， and Heilongjiang， all originating from the northern single-cropping zone. Cluster III consisted of Dingshu No. 4， characterized by purple-red tuber eyes. Cluster IV included two introduced materials. The two clustering results showed consistency in the grouping but differences in the classification of subgroups. 17 materials were classified into the same group under both clustering conditions. Combining phenotypic and SSR markers provides a more comprehensive and accurate reflection of the genetic differences among the germplasm resources.