Moroccan chestnut populations cover a small area in northern Morocco. They occupy great environmental and socio-economic interests. Nonetheless, such interests remained unvalued and little studied. This study aims to characterize and assess phenotypic similarities (or disparities) among the Moroccan populations of C. sativa via univariate and multivariate analyses of leaves traits. The experiments involved 6200 leaves from 31 populations, 10 trees per population, and 20 leaves per tree. Six morphometric parameters [blade area (S), perimeter (P), lamina length (LL), lamina width (LW), distance between the base and largest width (DBW), petiole length (PL)] and four ratios [LL/LW, LL/LP, LL/DBW, DBW/LP] were analyzed. Analysis of descriptive statistics within and between populations showed large variations among the studied parameters. This trend was supported as well by the analysis of variance (ANOVA) which revealed a highly significant differences (P <0.0001). Furthermore, the statistical analysis revealed a high phenotypic plasticity. The correlation analysis showed a significant relationship between most of the studied parameters. The principal component analysis (PCA) assigned the studied populations into two groups. We conclude that the studied parameters have great potential to determine the phenotypic polymorphism of C. sativa populations