Our lives are dominated by a complex visual world, and spatially selective attention allows us to process only the most relevant information. Previous evidence suggests that if possible locations of stimulus presentation are delineated, attention affects processing in a spatially graded manner. This gradient is seen in both behavioral measures and in visual evoked potentials (VEPs). Stimuli presented close to cued regions elicit faster responses and larger VEPs than those presented farther away. However, both position in the visual field and allocation of attention may contribute to the observed gradients. These relative contributions can be distinguished by comparing responses on physically identical trials when attention is directed to locations at various distances from the stimuli. In the current study, participants attended to one of 12 squares arranged in a circle around fixation. Letters appeared individually, each in one of the squares; 80% were O’s (standards) and 20% were X’s (deviants). Participants were instructed to press a button when an X appeared at the attended location. The largest amplitude N1s (150-200 ms) were observed when participants attended to the location where a standard was presented. VEPs elicited by standards showed evidence of asymmetric attentional gradients. Specifically, the gradient of facilitation spread down more than up. Results also showed that attention had differential effects on the stages of processing indexed at specific time windows. These results confirm that attention can be applied to visual processing in a spatial gradient, reveal its asymmetric distribution, and elaborate on the timing of its selectivity.