An experiment was conducted at three WI locations to determine short and long-term impacts of different seeding rates for kura clover (Trifolium ambiguum M. Bieb.). Four kura clover lines (Rhizo, Cossack, Endura, and WISC1) were sown in mid-April 1999 and weeds were controlled with herbicides over the next 4 years; during the establishment year, yields were not measured. In 2000 and in subsequent years, no yield differences were detected between seeding rates of 8 and 16 kg/ha. By 2001, yield from stands seeded at 4 kg/ha were not different from the higher seeding rates, and by 2002 yield from plots sown at 2 kg/ha were within 20% of those sown at higher rates. Rhizomes allowed kura clover to spread to open areas when weeds were suppressed. In dry environments, stands originally sown at higher rates performed relatively better than those sown at lower rates, perhaps because of a greater number of taproots per square meter. Endura, Cossack, and WISC1 performed similarly but Rhizo yielded significantly less than the others at all seeding rates in all environments. Kura clover has a remarkable ability to compensate for initial thin stands by spreading via rhizomes.