The Intermountain West is dominated by big sagebrush communities (Artemisia tridentata subspecies) that provide habitat and forage for wildlife, prevent erosion, and are economically important to recreation and livestock industries. The two most prominent subspecies of big sagebrush in this region are Wyoming big sagebrush (A. t. ssp. wyomingensis) and mountain big sagebrush (A. t. ssp. vaseyana). Increased understanding of seed bank dynamics will assist with sustainable management and persistence of sagebrush communities. For example, mountain big sagebrush may be subjected to shorter fire return intervals and prescribed fire is a tool used often to rejuvenate stands and reduce tree (Juniperus sp. or Pinus sp.) encroachment into these communities. A persistent seed bank for mountain big sagebrush would be advantageous under these circumstances. Laboratory germination trials indicate that seed dormancy in big sagebrush may be habitat-specific, with collections from colder sites being more dormant. Our objective was to investigate seed longevity of both subspecies by evaluating viability of seeds in the field with a seed retrieval experiment and sampling for seeds in situ. We chose six study sites for each subspecies. These sites were dispersed across eastern Oregon, southern Idaho, northwestern Utah, and eastern Nevada. Ninety-six polyester mesh bags, each containing 100 seeds of a subspecies, were placed at each site during November 2006. Seed bags were placed in three locations: (1) at the soil surface above litter, (2) on the soil surface beneath litter, and (3) 3 cm below the soil surface to determine whether dormancy is affected by continued darkness or environmental conditions. Subsets of seeds were examined in April and November in both 2007 and 2008 to determine seed viability dynamics. Seed bank samples were taken at each site, separated into litter and soil fractions, and assessed for number of germinable seeds in a greenhouse. Community composition data from each site, as well as several environmental variables, were used to evaluate seed viability within the context of habitat variation. Initial viability of seeds used in the seed retrieval experiment was 81 and 92 percent for mountain and Wyoming big sagebrush, respectively. After remaining in the field for 24 months, buried Wyoming big sagebrush seeds retained 28-58 percent viability,11-23 percent of seeds under litter remained viable, and no seeds remained viable on the surface (estimates are 95-percent confidence intervals). The odds of remaining viable did not change from 12 to 24 months. However, after 24 months the odds of seeds beneath litter being viable decreased to 75 percent of the odds of viability at 12 months. Similar to Wyoming big sagebrush, buried seeds of mountain big sagebrush were 31-68 percent viable, seeds under litter retained 10-22 percent of their viability, and no surface seeds were viable after 24 months. Both subspecies of big sagebrush had some portion of seed that remained viable for more than one growing season provided they were buried or under litter. Although seeds beneath litter may remain viable in intact communities, seeds are susceptible to incineration during fires. Nine months after seed dispersal, seed bank estimates for Wyoming big sagebrush ranged from 19 to 49 viable seeds/m2 in litter samples and 19-57 viable seeds/m2 in soil samples (95-percent confidence interval). For mountain big sagebrush, estimates were 27-75 viable seeds/m2 in litter samples and 54-139 viable seeds/m2 in soil (95-percent confidence interval). The number of viable seeds present in the seed bank 9 months after seed dispersal was not significantly different from the number present immediately after seed dispersal. Seed viability was highest in mountain big sagebrush sites for seeds on the surface and beneath litter, but decreased after one season. Buried seeds of both subspecies were in equal abundances and may be insulated from the effect