Changes in dormancy and viability of freshly harvested seed of yellow nutsedge buried in the field, at depths of 5, 25, and 50 mm, were studied over a 2-yr period. The number of dormant, viable seed decreased most rapidly at 5 mm because more seed lost their dormancy and germinated (up to 30%) than at lower depths. Loss of viability through decay of seed appeared similar at all depths. No seedlings emerged from any depth. In a laboratory experiment, conducted in optimal soil moisture conditions, seedlings emerged from all seed that germinated, demonstrating that adequate soil moisture was critical for seedling establishment. This was substantiated in a second field experiment in which irrigation and mulching greatly increased seedling survival. The number of seedlings surviving in irrigated plots was 0.78% of seed sown and 0.03% in the rain-fed plots. The role of seed in the establishment of yellow nutsedge infestations is probably of little importance in dryland cropping areas despite the longevity and viability of the seed.