Sexual reproduction is known to be an important means of propagation for native Japanese knotweed populations in Asia. For naturalized populations in the United States, however, its relative importance compared with propagation by rhizome and stem fragments has not been established. This article presents two related studies that address two basic questions regarding the potential for propagation by seed: (1) are seeds that are produced by Japanese knotweed in areas of Philadelphia, PA, capable of germinating in the field? If so, (2) when and how rapidly during the year do these seeds acquire germinability? Field germination experiments assessed germination of planted seeds at intervals during the growing season at two different study sites (one in 2000 and the other in 2002) and showed that germination was common at both sites. The 2002 experiment also demonstrated that germination occurred both within existing stands of Japanese knotweed and in areas well removed from existing stands, and that both planted and naturally occurring seeds germinated. Experiments on seasonal acquisition of seed germinability used batches of seeds collected weekly between early September and November 2000 from three different study sites. Results showed that at two of the study sites, a rapid increase in germinability from less than 10% to greater than 90% occurred between collection dates in mid-September and mid-October 2000. A smaller and more gradual change occurred at the third site, where maximum germinability was roughly 50%. Regarding management of Japanese knotweed, the field germination experiments suggest that spread of this plant by seed is a realistic possibility in the United States. Because of the rapidity with which germinability is acquired, the high germinability attained, and the large number of seeds produced, we recommend that female (male-sterile) plants or their inflorescences be removed from seed-producing populations before the formation of fruits to minimize spread by seed.