Previous phylogenetic work investigating the origins of the grain amaranths is plagued by two methodological faults. Primarily, no study to date has adequately sampled all of the grain species as well as the two putative weedy progenitors across their entire native species’ ranges. Second, no study to date adequately investigates the genetic diversity component of both putative weedy species (Amaranthus hybridus L. and Amaranthus quitensis Kunth). Two hundred fifty‐eight individuals from 56 taxa representing the three grain amaranth species and two putative weedy progenitors were sampled from their native species’ ranges and subjected to microsatellite testing. Data were analyzed using principal component analysis, clustering algorithms, and genetic distance neighbor‐joining tree dendrograms. Standard genetic diversity measures indicate strong inbreeding coefficients in all studied taxa. Principal component analysis, clustering algorithms, private alleles, and tree dendrograms all resolve a distinct A. quitensis group. The grain amaranths and A. hybridus were found to group together. Results from 11 simple sequence repeat loci indicate the following: (i) A. hybridus is the progenitor of all three grain amaranths in at least two separate domestication events, (ii) A. quitensis is its own species and not a variant of A. hybridus, and (iii) Amaranthus caudatus L. and Amaranthus hypochondriacus L. appear to be closely related.

Amaranth (Amaranthus caudatus L.) is an important seed crop in the Andes. It has excellent nutritional value and ample capacity for growth under drought, heat, and soil nutrient deficiency. The objective of this study was to characterize and quantify genetic diversity among a series of 178 mostly Peruvian amaranth individuals genotyped from among 48 accessions using 96 biallelic single nucleotide polymorphism (SNP) markers in KASPar assays on a 96.96 Fluidigm EP1 platform. The 178 seed genotypes included the predominant white‐opaque (waxy), white‐vitreous (nonwaxy), black, and brown types, all from within the Centro de Investigacion de Cultivos Andinos‐Universidad Nacional de San Antonio Abad del Cusco collection. The samples separated into two groups based on the seed color parameter (black, brown, and white‐vitreous versus white‐opaque) with STRUCTURE analysis but did not form distinct genotypic groups based on geographic origin. When analyzed as seed‐color subgroups—black vs. brown vs. white‐vitreous vs. white‐opaque—83.46% of the genetic variation was explained by genetic differences within subgroup. The highest average observed heterozygosity (Ho) within color‐based subgroups was 0.2612 in the vitreous‐seeded subgroup, and within‐group expected heterozygosity (He) was highest in the brown‐seeded subgroup (0.398). Cluster analysis and principal component analysis partitioned the amaranth accessions into two clusters with four distinct subclusters. Analyses also revealed evidence for selection of nonwaxy, vitreous‐seeded amaranth in the Andes.