Grazers and granivores have the potential to affect seed banks. Several studies have examined the impact of these herbivores on the aboveground vegetation, but few have looked at how they influence the seed bank. I asked whether both grazers and granivores alter the seed bank at the Sevilleta National Wildlife Refuge. Long-term experimental plots were installed in 1996 to exclude grazers and granivores from a grassland and shrubland. Soil samples were collected from these plots and seeds were germinated in a greenhouse. The grassland had significantly more species in its seed bank than the shrubland. Also, the seed bank composition differed significantly between the two sites. However, the number of species in the seed bank did not vary among herbivore treatments nor did total seed numbers vary among treatments at the grassland. At the shrubland, in contrast, plots that excluded both herbivores had fewer total seeds than control plots and plots where only grazers were excluded. Therefore, although herbivores play some role in the shrubland, herbivores do not reduce seed numbers at either site. Thus, seed bank size is not controlled by the consumption of seeds from these herbivores, but by some other factor (e.g. disturbance or abiotic events).