I used an individual-based model to simulate population dynamics and changes in genetic variation of horses in the Pryor Mountain Wild Horse Range under a range of management strategies. Management strategies included population objectives of 90 to 180 horses and removal and/or contraception of animals of random, young, or old ages. Populations controlled only by removals fluctuated widely in size and generally exceeded objectives by >20%. Populations controlled by contraceptives remained within 3% of objective. Over a 200-year period, populations were very likely to lose alleles with an initial frequency of <3%, which included nearly 30% of alleles in the population. Allelic diversity was lost faster than population-level heterozygosity. Removal or contraceptive treatment of older animals hastened the loss of genetic variation compared to treatment of random-aged or young animals. Rate of loss of genetic variation was directly related to generation interval, and differences in loss of genetic variation were greater between treatments of old and young animals than between populations that differed in size by a factor of 2. These results apply to management of many long-lived species in harvested populations and they illustrate the importance of actions that modify age structure and generation interval.