One hundred and thirty six peanut cultivars (including main breeding parents)originating in different provinces in China were selected as research population and the quality traits such as protein content、oil content、oleic acid content were tested. Whole genome of the cultivars were scanned with 64 SSRs of a high level of diversity and located dispersedly on a composite genetic linkage map constructed at HAAS. By using Structure2.3.1, the genotypes of peanut were used to detected the cultivars and to assign individuals to subpopulations by Q. Then by using Tassel 2.1, GLM (Q) model was successfully established by joining 136 genotypes of germplasm and quality traits in 3 years with Q of population structure as the covariate. The main results shows: ①136 cultivars according to individual genotype was assigned to 5 subpopulations. Because the population structure was consistent with the characters and geographic origins of cultivars respectively, these selected germplasm was proved to be suitable for association analysis as a result.②By constructing GLM(Q) with 64 genome-wide SSR loci and quality traits from the year 2010 to 2012, 18、31、26 SSR markers associated with quality traits significantly were detected respectively.③Integrating the information of SSR associated with quality traits from 2010 to 2012 , 4 SSR could be identified repetitively, and 40 alleles corresponding to these markers were identified totally.