Precipitation and net primary productivity are positively correlated in arid environments. Both variables are, in turn, correlated with mammal species richness, but this relationship is not necessarily positive. With increasing precipitation in arid areas of low to moderate productivity, mammal richness increases linearly; as rainfall and productivity increase beyond this point, mammal richness is known to decline in some areas, producing a relationship that has been termed “unimodal” or “humped.” In the Great Basin of the arid western United States, studies of the relationship between rodent species richness and precipitation have revealed only a positive relationship between these two variables. It has, however, been argued that if areas of higher precipitation were to be sampled within this region, the decline phase would become evident. When latest Pleistocene and Holocene small mammal assemblages from the northern Bonneville Basin (central Utah) are examined across a temporal moisture gradient, species richness declines as moisture declines. Since the Great Basin was significantly moister during the latest Pleistocene and Early Holocene than it has been since that time, the unimodal response model does not appear to apply to this region.