Recently, a number of papers have addressed the use of pedigrees in the study of wild populations, highlighting the value of pedigrees in conservation management. We used pedigrees to study the horses (Equus caballus) of Assateague Island National Seashore, Maryland, USA, one of a small number of free-ranging animal populations that have been the subject of long-term studies. This population grew from 28 in 1968 to 175 in 2001, causing negative impacts on the island ecosystem. To minimize these effects, an immunocontraception program was instituted, and horse numbers are slowly decreasing. However, there is concern that this program may negatively affect the genetic health of the herd. We found that although mitochondrial DNA diversity is low, nuclear diversity is comparable to that of established breeds. Using genetic data, we verified and amended maternal pedigrees that had been primarily based on behavioral data and inferred paternity using genetic data along with National Park Service records of the historic ranges of males. The resulting pedigrees enabled us to examine demography, founder contributions, rates of inbreeding and loss of diversity over recent generations, as well as the level of kinship among horses. We then evaluated the strategy of removing individuals (using nonlethal means) with the highest mean kinship values. Although the removal strategy increased the retained diversity of founders and decreased average kinship between individuals, it disproportionately impacted sizes of the youngest age classes. Our results suggest that a combined strategy of controlled breeding and immunocontraception would be more effective than removing individuals with high mean kinships in preserving the long-term health and viability of the herd.