Abstract: The bay-Asturcón pony (A21) population recovery started recently. A total of 297 samples (180 from founder individuals) belonging to the A21 population were genotyped for: a) a set of 15 microsatellites; and b) the presence of the two recessive chestnut alleles reported in the literature (e and ea) on the MC1r gene (locus Extension) and the deletion on the ASIP gene (locus Agouti) associated with recessive black coat in horses. The extent to which the genetic variability of the A21 population could be affected by selection strategies aimed at decreasing the frequency of the chestnut allele was quantified in terms of gene diversity and allelic richness. The possible genetic impact of a controlled introgression of A21 offspring into the black-coated Asturcón (A20) population was also assessed using 261 available A20 genotypes. The wild alleles for the Extension and Agouti loci (E and A, respectively) were the most frequent (77.8% and 59.4%) in the A21 population. Both the e and ea recessive chestnut alleles were identified with frequencies of 21.2% and 1.0%. As expected, the contribution to overall diversity of the founder subpopulation was always higher than that of the A21 offspring subpopulation. Total contribution of the offspring subpopulation to overall allelic richness was negative (- 1.84%), showing that all their alleles were present in the founder subpopulation. Although favourable, the chestnut carrier individuals had poor contributions to overall gene diversity and tended to have negative contributions to allelic richness. The elimination of the chestnut carrier A21 individuals would not affect genetic variability to a significant extent. Therefore, efforts for preserving the genetic variability in the A21 population could focus on a careful planning of matings between individuals free of the recessive chestnut alleles. The expected influence of a controlled introgression of A21 offspring on the genetic variability of the A20 population was always favourable. The gene diversity of the introgressed population was higher than that of the original A20 population for both gene diversity (GDT = -1.2% ± 0.04%) and, particularly, allelic richness (CT = 4.9% ± 0.27%). Very limited gene flow may increase the number of alleles of the A20 population but also would yield a balance of the allelic frequencies at a population level. Therefore, a common breeding policy for the two Asturcón pony populations may ensure the viability of this prominent genetic resource.