The Latin American shrimp Litopenaeus stylirostris was introduced in three different Pacific islands (Tahiti, New Caledonia via Tahiti, and Hawaii) and hatchery-propagated for 7–25 generations to develop shrimp farming based on these domesticated stocks. Three microsatellite markers have been used in an attempt to assess the genetic bases of the populations available to start a selective breeding program. The comparison of eight hatchery stocks (five New Caledonian, two Hawaiian and one Tahitian stocks) and one wild Ecuadorian population showed a much lower variability in the domesticated stocks than in the wild population, especially in New Caledonia and Tahiti (2–3.7 vs. 14–27 alleles per locus; 20–60% vs. 90% expected heterozygosity). The Tahitian and the New Caledonian stocks share the same alleles, suggesting that the loss of alleles occurred during the common past of these populations. On the contrary, New Caledonian and Hawaiian populations share only one common allele at the three loci studied. Although the low genetic variability and the resulting inbreeding of the New Caledonian stocks do not seem to affect their present performance, the results of this study demonstrate the usefulness of the introduction of new stocks in order to increase the potential responses to new controlled or uncontrolled selective pressures. The introduction in New Caledonia of the Hawaiian domesticated stocks, which would provide the local shrimp industry with 40% of the allelic diversity of the species, is advised and preferred to the one of wild animals in order to take advantage (i) of the spontaneous selection which occurred during domestication and (ii) of their favourable sanitary "specific pathogen free" status (no presence of four viruses: WSV, YHV, IHHNV, TSV) which limits the risk of introduction of pathogens.