Three experiments explored the effects of global and local processing on the distribution of spatial attention. In experiment 1, participants were asked to discriminate either a large or a small circle presented right after the display of the compound letter whose identity should be later reported. The result was that an attended level of the compound letter affected the size judgment of the target circle. Experiment 2 made participants detect the position of gap in a small circle to investigate whether spatial arrangement of compound letters would affect the specific location of spatial attention, with no convincing result. By presenting all the four small circles at once only one of which has the gap, experiment 3 showed that more spatial attention would be given to the vertex position of the compound letters than to the other positions. These results suggest that both global and local processing influence the size of spatial attention, and make shift local spatial attention to the informative area of the compound letter as well.