Using the example of a reintroduced bighorn sheep population in Badlands National Park, South Dakota we demonstrate the usefulness of neutrality tests and demographic data for detecting a severe genetic bottleneck (Ne < 10). From demographic data the effective population size of the founding population at Badlands was estimated to be six, and a heterozygosity excess test revealed evidence of a severe population bottleneck. We discuss the criteria for intervention when there is evidence of a severe bottleneck, and propose methods of mitigating the potentially deleterious long-term consequences of such bottlenecks. These issues are presented in the context of bighorn sheep reintroductions, but the issues are also of general importance to restoration efforts involving other large vertebrates.