Over time, farmers have always made chestnut selections by multiplying the most interesting plants. Indeed, a good knowledge of the phenotypic and genotypic variability of ecotypes is the basis of plant breeding. However, Moroccan chestnut ecotypes are neither classified nor characterized. This study aims to evaluate the genetic resources of Castanea sativa MILL. in northern Morocco through the morphometric characterization of leaves. Thirty-one villages in three regions were involved, ten trees per village and twenty leaves per tree were sampled. Then ten morphometric parameters were analyzed: lamina length (LL), lamina width (LW), petiole length (PL), distance from the base of the leaf to the widest point of the leaf (DBW), area (S), perimeter (P), and ratios LL/LW, LL/LP, LL/DBW, and DBW/LP. Analysis of the descriptive statistics within and between ecotypes initially showed a large variation in the ten parameters studied. This finding was supported by analysis of variance (ANOVA) which revealed a very highly significant difference (P <0.0001) for all parameters. Indeed, the analysis of agglomerative hierarchical clustering (AHC) made it possible to group the studied populations into 3 distinct groups based on the leaf area. Thus, the wide variability of the morphometric parameters of the leaves constitutes an important group of traits useful for conventional breeding programs, providing a knowledge base for the exploitation of local genetic resources and the in situ conservation of this heritage.