We investigated the role of foraging and escape substrates on foraging decisions of a Namib Desert gerbil, Gerbillurus tytonis (Rodentia), in the field. We used the giving-up density technique, which uses artificial seed trays to assess foraging decisions. We tested whether the foraging decisions made in the trays were an artifact of prior knowledge of a food source by comparing giving-up densities in trays that were randomly placed in different parts each night with that in trays that remained in fixed position throughout the experiment. There was no significant difference in giving-up densities between these two treatments, which suggests that prior knowledge of food location does not affect foraging decisions. We examined the preferences of gerbils for the different substrates using seed trays. The gerbils showed significant differences among habitats in the amount of seed taken and between new-moon and full-moon nights. They removed more seeds where the difference between the size of the seeds and the size of the sand substrate was large. This finding is congruent with the hypothesis that tactile discriminatory ability affects the foraging behavior of these rodents. Although these rodents alter their foraging behavior in response to predation risk, there was no significant difference in the amount of seeds taken in habitats differing only in the degree of compaction of escape substrates. Keywords: foraging behavior, predation risk, desert rodents, optimal foraging, Gerbillurus.