IntroductionIn Brazil, the plant group popularly known as “pedra‐ume‐caá” is used in folk medicine for the treatment of diabetes, and its raw material is commonly sold.ObjectiveThe aim of the study was to apply a method for chemical identification of extracts of dry pedra‐ume‐caá leaves using HPLC–high‐resolution mass spectrometry (HRMS) and NMR and develop a multivariate model with NMR data to authenticate commercial samples. In addition, to evaluate the biological activities of the extracts.Materials and MethodsDry extracts of Myrcia multiflora, Myrcia amazonica, Myrcia guianensis, Myrcia sylvatica, Eugenia punicifolia leaves, and 15 commercial samples (sold in Manaus and Belém, Brazil) were prepared by infusion. All the extracts were analysed using HPLC–high‐resolution mass spectrometry (HRMS), NMR, principal component analysis (PCA), and hierarchical cluster analysis (HCA). The antidiabetic effect of extracts was evaluated according to enzymatic inhibition. Their content of total phenols, cell viability, and antioxidant and antiglycation activities were also determined.ResultsHPLC‐HRMS and NMR analysis of these extracts permitted the identification of 17 compounds. 1H NMR data combined with multivariate analyses allowed us to conclude that catechin, myricitrin, quercitrin, and gallic and quinic acids are the main chemical markers of pedra‐ume‐caá species. These markers were identified in 15 commercial samples of pedra‐ume‐caá. Additionally, only the extracts of M. multiflora and E. punicifolia inhibited α‐glucosidase. All the extracts inhibited the formation of advanced glycation end products (AGEs) and showed free‐radical‐scavenging activity. These extracts did not present cytotoxicity.ConclusionThis study revealed the chemical markers of matrices, and it was possible to differentiate the materials marketed as pedra‐ume‐caá. Moreover, this study corroborates the potential of these species for treating diabetes.