Predation greatly influences many aspects of the ecology of desert rodents, from foraging behaviour to mechanisms of species coexistence to the evolution of specialised morphologies. Using a foraging.theory approach, we examine consequences of predation for assemblages of desert rodents from North America and the Middle East. In particular, we review experimental evidence that examines the influence of predation on foraging costs and foraging behaviour, explore how predation can act to structure communities, and discuss the role that predation may have played in the evolution of bipedal locomotion. Finally, we compare the importance of predation for the evolution of anti.predator behaviours and morphology, for population dynamics, and for community processes, with its magnitude and heterogeneity. In regard to foraging behaviour, desert rodents treat the risk of predation as a cost of foraging. They combine assessments of food and safety to arrive at foraging decisions, exploiting resource patches less intensively in response to increased predatory risk. The cost of predation can be up to 91% of the foraging costs of desert rodents, but the proportion is greater for Middle Eastern rodents than for North American rodents. In regard to community structure, predation can provide the niche axis as well as the necessary tradeoff for species coexistence. Despite the importance of predation in shaping the foraging behaviour of desert rodents, predation may not always influence species coexistence. Predation contributes to species coexistence at sites in the Sonoran and Great Basin deserts. But in the Negev Desert, where predation costs are the greatest, predation does not provide a mechanism of species coexistence. In regard to bipedal locomotion, predation most likely confers superior ability to avoid predators by improving sprint speed and ability to take evasive action, but at the expense of foraging ability in safe microhabitats. The evolution of bipedality will be favoured by situations where the risk of predation is great: the open microhabitat is riskier than the bush, the richest patches are found in the riskiest places, and rich patches are far apart. The magnitude of predatory risk will affect the evolution of anti.predator behaviour and morphologies. However, the importance of predation in community processes is not determined by its magnitude, but by its heterogeneity in time and space relative to the abilities of potentially coexisting species.