This study investigated the influence of spatial cueing (valid/invalid/no cue) on visual discrimination in human and non-human primates. We employed a spatial resolution task which required the accurate discrimination of the orientation of a Landolt "C" ring. The C appeared as single target in specific retinal locations while subjects maintained fixation of a central fixation point. The minimal discernable size of the "C" (=acuity threshold) was determined as a function of cue condition, retinal eccentricity (3 degrees -15 degrees ), and stimulus onset asynchrony (SOA) (200-1100 ms). For both species, we found consistent benefits from spatial cueing with differences in absolute thresholds ranging from 6% to 25%. These differences increased with retinal eccentricity and decreased with longer SOAs. Further experiments performed with humans only, showed that the effect of spatial cueing on visual discrimination is independent of spatial uncertainty, i.e. the number of possible target locations (2 versus 4), but fades with longer target presentation times. From our results we draw the following conclusions. (i) Since sensory noise and spatial uncertainty was small in our tasks, spatial shifts of attention involve signal enhancement in both, human and non-human primates. (ii) The similarity of the results obtained for humans and macaque monkeys indicates that the latter may serve as a suitable model system in studies trying to tackle the neural underpinnings of attentional control. (iii) In order to elicit robust effects on visual discrimination by spatial shifts of attention, a paradigm comprising short SOAs (approximately 200 ms) and target presentation times (approximately 150 ms), and retinal eccentricities larger than approximately 9 degrees seems most promising.