The purpose of this study was to explore the phenotypic diversity of Castanea mollissima Bl. in Yanshan Mountains, and evaluate the nut appearance and economic traits targetedly using a variety of analytical methods. The phenotypic datasets (11 descriptive traits and 15 numerical traits) in 118 chestnut variety resources in Yanshan Mountains were collected. The Shannon-Wiener diversity index, principal component analysis, membership function and cluster analysis were used to evaluate and guide the appropriate utilization methods according to the nut phenotypic characteristics. The value of Shannon-Wiener diversity index（H'）of 11 descriptive traits ranged from 0.33 (number of pubescence) to 1.14 (color of nut), and the value of coefficient of variation was 11.63 % (obviousness of stripes) - 51.07 % (wiring shape). The value of coefficient of variation of 15 numerical traits ranged from 3.09% (kernel rate) to 24.69% (shell weight), and the value of coefficient of variation of shell weight, bar size, kernel weight and nut size was greater than 20.00 %. The Shannon-Wiener diversity index ranged from 1.85 (base width) to 2.09 (nut length). There were significant or extremely significant correlations among most numerical traits. The results of principal component analysis showed that the cumulative contribution rate of the first four principal components was 80.768 %, and contribution ratio of the first principal component (nut width, nut thickness, nut size, surface arc length and so on) was the largest, which could be used as a comprehensive index related to nut size. The variety resources with higher comprehensive scores of nut processing economic traits included Jianli, Nanchui 5, Xinglong 1, Maxingli, Yanli 1, Dabanhong, Yanli, Pingquan (Ge), Funingbaopi, Timashisheng.