The objective of this study was to assess genetic diversity among 6 different wild ginseng populations from New York, Kentucky, North Carolina, Pennsylvania, Tennessee and Virginia, and to compare these wild populations to one cultivated population. RAPD markers were used to estimate the genetic difference among samples from the 7 populations. The 64 random primers were screened, and 15 primers were selected which exhibited the 124 highly reproducible polymorphic markers. The ratio of discordant bands to total bands scored was used to estimate the genetic distance within and among populations. Multidimensional scaling (MDS) of the relation matrix showed distinctive separation between wild and cultivated populations. The MDS result was confirmed using pooled chi-square tests for fragment homogeneity. This study suggests that RAPD markers can be used as population-specific markers for American ginseng.