Longan (Dimocarpus longan Lour.) is an important underutilized but nutritionally rich fruit species in India, rich in vitamins, nutrients, and flavonoids. It lacks diversity and distribution studies and characterization of cultivars aiming at genetic improvement. This study aimed to explore the morphological and physiochemical diversity among 20 longan genotypes grown under subtropical Northern India, during 2019-2020. A total of 15 qualitative and 34 quantitative traits were assessed. The Spearman's rank correlation and Pearson's correlation coefficients were used for qualitative and quantitative traits, respectively. Significant correlation coefficients were recorded among all the studied quantitative traits. Path analysis was employed to determine the cause-and-effect relationship among the studied traits. Results revealed that aril/seed weight ratio, followed by plant height, petiole length, fruit width, ascorbic acid, fruit length/width ratio, titratable acidity, TSS/acidity ratio, and duration of panicle initiation to flowering showed a direct positive effect towards the number of fruits per bunch. Furthermore, PCA was done separately for the 15 qualitative and 34 quantitative traits. The findings revealed that the first ten major principal components (PCs) explained 92.99% of the total variation in quantitative traits. In qualitative traits, the first five major components accounted for 72.19% of the total variation. The bi-plot between PC1 and PC2 identified that genotypes LGC-6, LGC-47, LGC-37, LGC-1, and LGC-18 for quantitative traits and LGC-6, LGC-18, LGC-37, LGC-61, and LGC-59 for qualitative traits were as very diverse, suggesting their suitability as parental material in breeding programs. Agglomerative hierarchical clustering (Ward's method) grouped the genotypes into two main clusters, based on quantitative and qualitative traits. These clusters can represent distinct gene pools and can be effectively utilized in hybridization programs to enhance genetic improvement in longan.