In this study, 39 specimens belonging to Lespedeza species (Lespedeza cyrtobotrya, L. bicolor, L. maximowiczii, and Lespedeza cuneata) (Leguminosae) were classified phenotypically and genotypically. We constructed a phylogenetic tree based on the combined nrDNA (internal transcribed spacer; ITS) and cpDNA (trnL-trnF) sequences with the aim of classifying the genotypes. Samples were mainly divided into three genotypes. Samples of L. cyrtobotrya and L. bicolor were mixed in a single branch, whereas samples of L. maximowiczii and L. cuneata were clustered within species, respectively. We performed a liquid chromatography-electrospray ionization-mass spectrometry-based metabolite profiling analysis to classify the phenotypes. Multivariate statistical analyses such as principal component analysis (PCA) and hierarchical clustering analysis (HCA) were used for the clustering pattern analysis and distance analysis between species, respectively. According to the PCA and HCA results, leaves were classified into four phenotypes according to species. In both the genetic and chemotaxonomic classification methods, the distance between L. cyrtobotrya and L. bicolor was the closest between species, and L. cuneata was the farthest away from the other three species. Additionally, orthogonal partial least squares-discriminant analysis was employed to identify significantly different phytochemicals between species. We classified L. cyrtobotrya and L. bicolor by identifying significantly different phytochemicals. Interestingly, leaves and stems showed different phenotypic classifications based on the chemotaxonomic classification. Stem samples of the other three species were mixed regardless of species, whereas L. cyrtobotrya stem samples were clustered within species. The phenotypic classification of leaves coincided more with the genotypic classification than that of stems. Key message We classified four wild-type Lespedeza sp. by analyzing the combined nrDNA (ITS) and cpDNA (trnL-trnF) sequences. We also classified leaves and stems of Lespedeza sp. by applying liquid chromatography-mass spectroscopy-based metabolite profiling.