The formation of fish breeding stocks for fish farming or conservation programs is commonly carried out from the capture of fish in natural environments. Information on the geographic and genetic origin of these stocks is important to guide actions that allow correct management in captivity and, when lost, harm production and genetic conservation. In this sense, the objective of this study was to evaluate the genetic diversity and origin of two breeding stocks of matrinxã, Brycon amazonicus (INPA, Amazonas - INPA and Nova Motum, Mato Grosso - NM). A total of 68 caudal fin samples were collected, including 33 INPA samples and 35 NM samples. Twenty pairs of microsatellite primers were tested, but only seven primers showed satisfactory amplification, amplifying 41 alleles ranging from 187-318 bp. The polymorphic information content ranged from 0.135 (Borg25) to 0.782 (Bh6). Exclusive alleles were observed for both populations (INPA: 04 and NM: 18). Allelic richness results revealed that there was increased loss of genetic variation in NM, indicating a lower evolutionary potential of this stock. The average values of the observed heterozygosity corroborated this statement; however, there were high values for INPA (0.545) and NM (0.475), signifying an adequate genetic variability. An imbalance was found in the Hardy-Weinberg equilibrium at the Borg59 locus in INPA (P < 0.05), possibly due to the effect of null alleles, but was attributed to a founder effect. For NM, an imbalance in the Hardy-Weinberg equilibrium was observed at loci BoM13 and Bh6, which together with the results of the mean inbreeding coefficient values demonstrated the presence of genetic drift. The analysis of molecular variance showed greater variation within populations than between them, and was confirmed by the genetic differentiation value (0.086 - moderate genetic differentiation) and by the distance and genetic identity values (0.273 and 0.761, respectively). Bayesian analysis designated a value of K = 2, with the presence of structuring for NM and INPA; however, with correlated allelic frequencies, confirming a common origin. This origin was corroborated by the presence of gene flow through the number of migrants (5.691). Based on these results, there was a moderate genetic variability for INPA and NM and their common origin was confirmed. Recommendations are also included to minimize the probability of inbreeding processes or genetic drift in the studied stocks.