Covert orientation of attention was studied in 30 adults who fixated warning cues and pressed a button at target onset. Directional cues (arrows) indicated the most probable (p = 0.8) side of target occurrence. Subjects responded fastest when validly cued, slowest to invalidly cued targets, and at an intermediate rate when the cue (a cross) was not directional. Directional cues took longer to evaluate (increased N1 and P2 latencies) and produced more focussed attention and greater response preparation (enhanced CNV and P3 amplitude) than non-directional cues. These findings indicate that the expectancy of a target can be manipulated by a spatial cue at three levels, sensory, attention, and response preparation, and lead to changes in the sensory perceptual processing of the target. Validly cued targets produced an increase in P1 amplitude reflecting attention enhanced sensory processing whereas invalidly cued targets increased N1 and P3 amplitudes reflecting the re-orientation of attention, and further processing and updating of information required of low probability stimuli respectively. P3 latency to invalidly cued targets was also delayed reflecting the additional processes required to shift attention to a new location. The P3 latency validity effect was smaller than that found for response time suggesting response execution may also be affected by spatial attention.