Sustained harvest of wild North American ginseng (Panax quinquefolius L.) for over two centuries has led to heightened conservation concern and a recent interest in the population genetics of this species. This study examined allozyme variation from 32 wild and 12 cultivated populations of American ginseng to: (1) document the amount and distribution of genetic variation over a wide portion of the species' natural range, (2) examine genetic differences between wild and cultivated populations, and (3) provide indirect estimates of its breeding system. Strong genetic differences between wild and cultivated populations were found in the amount of variation within populations and the distribution of variation among populations. Wild populations were significantly lower in all within-population diversity measures, but contained significantly higher levels of variation partitioned among populations. Similarities between wild and cultivated populations were also found. As a whole, cultivated and wild groups shared nearly all alleles, and populations of both groups showed strong homozygote excess compared to expectations under random mating. The homozygote excess is best explained by high levels of selfing. In wild populations, significant correlations were found between genetic diversity and estimated population size, and between interpopulation genetic distance and geographic distance. Overall, the results for wild populations suggest that they are influenced by high levels of genetic drift and low migration among populations. Conservation implications for American ginseng are discussed with particular emphases on: (1) the current debate surrounding the existence of wild populations, (2) the ongoing practice of introducing cultivated seed into wild populations, and (3) the collection of genetic material for the establishment of breeding programs.