The 2-year data of 5 quality traits and 5 yield traits of 366 long-staple upland cotton germplasm resources were genetically analysed using genotype and gene-environment interaction genetic models, and genetic diversity analysis of phenotypes and correlation analysis, cluster analysis, principal component analysis and multiple linear regression analysis of genotypic values were performed to clarify the value of the use of various types of germplasm resources. The results showed that the coefficients of variation of 366 long-staple upland cotton varied from 1.34% to 11.80% for quality traits and 7.95% to 54.09% for yield traits in two years; the diversity indices of two-year quality traits ranged from 1.67 to 2.03, with an average of 1.879, and those of two-year yield traits ranged from 1.42 to 1.99, with an average of 1.782, and the average of 1.782, and the average of 1.879. The average genetic diversity was 1.782, indicating that the 366 germplasm resources were more different from each other and rich in germplasm types. Cluster analysis divided the 366 medium and long-staple upland cotton germplasm resources into five classes, Class 1 included 63 germplasm, which belonged to the class of materials with high boll weight, high lint index and good uniformity; Class 2 included 41 germplasm, which belonged to the high-yielding materials with high lint percentage, high number of bolls per plant and high lint yield per plant; Class 3 included 71 germplasm, which belonged to the class of materials with short lint length and low boll weight; Class 4 included 84 germplasm, which belonged to the class of materials with long lint length and low boll weight; Class 4 included 84 germplasm, which belonged to the class of materials with long lint length, high Markelin number, and high lint yield per plant. longer lint length, lower micronaire value and higher specific strength; Category 5 includes 107 germplasm, which belongs to high-quality materials with the longest lint length, the lowest Micronaire value and the highest specific strength. By calculating the composite score (F-value) for each category of germplasm resources, nine varieties were screened to be greater than the control Jingang (0.87), four varieties were between the control Jingang and the control Chinese Cotton -3 (0.80), and five materials were between the control Chinese Cotton -3 and the control Xinluzhong 87 (0.74). Multiple linear regression analyses of the main traits yielded two linear regression models for lint length and lint yield per plant. The 366 long-staple upland cotton germplasm resources are rich in genetic diversity; the 21 high-quality germplasm resources screened can be used as high-quality parents to improve the fibre quality of local upland cotton and provide reference for upland cotton germplasm innovation and genetic breeding.