Most varieties of the American cranberry (Vaccinium macrocarpon) cultivated today were selected from native selections or breeding progeny between the late 1800s and mid-1900s. We have previously shown using RAPDs that contamination, i.e., a mixture of genotypes, is common in commercial bogs. One source of contamination could be establishment of selfed progeny. The purpose of this study was to determine how effective RAPDs would be in distinguishing selfed progeny from the parent. Results suggest that the number of scorable polymorphic bands is low compared to outcrossed or unrelated progeny. Thus, five to nine primers were used as compared to the three primers normally required to separate outcrossed and unrelated clones. Segregation of some RAPD bands was not consistent with expected mendelian ratios. However, using 9 to 12 polymorphic bands, only 3% to 5% of the selfed progeny had fingerprints identical to the parent. Additional primers should further reduce this percentage. It was also noted that certain cultivars exhibited a large number of non-parental bands. The origin of the non-parental bands has not yet been determined.