Inhibition of return (IOR) refers to slower responses to stimuli appearing at previously attended locations than new locations. Although a large number of ERPs have found that there is a stable difference (Nd250) between the target evoked ERPs at the cue and uncued locations around 250-300 ms, the direction of this difference is not consistent. A possible reason for this inconsistency is that these studies used a different arrangement of stimuli. So far, few studies have directly investigated the impact of this factor on the expression of IOR on Nd250. In the current study, we combined ERPs with the cue-target paradigm and manipulated the arrangement of stimuli to examine whether it could affect the expression of IOR on Nd250 or not. Sixteen paid, naive volunteers recruited from the campus of Soochow University (10 females, 6 males, 20±2.37 years old) participated in the experiment. All participants had normal or corrected visual acuity and had no experience in similar experiments. All participants gave written informed consent according to the standard of the Declaration of Helsinki. The study was approved by the Academic Committee of the Department of Psychology, Soochow University. The current study used a 2 (arrangement of stimuli) × 2 (cue-target relationship) within-subject design. There are two levels of the arrangement of stimuli (horizontal and vertical arrangement) and two two levels of the cue-target relationship (cued and uncued). Under cued trials, the cue and the target were presented at the same location; under uncued trials, the cue and the target appeared at different locations. While the cue-target relationship factor was varied within blocks, the arrangement of stimuli was varied between blocks. The arrangement of stimuli order was balanced across the 12 experimental blocks with an ABBA-BAAB-ABBA design and counterbalanced across participants. The experiment was conducted in a quiet, dimly lit room. The participants were sat around 80 cm away from the screen and responded with a gamepad. During the experiment, the participants were required to make a discrimination task on targets (@ vs. &) occurred either at previous cued locations or uncued locations. EEG and EEG data were collected using a Neuroscan system with a 64-channel Quik-Cap. Although the effect of the arrangement of stimuli on the polarity of Nd250 was not observed, we do observe a significant effect of the arrangement of stimuli on the topographic distribution of early ERP components: while the P1 cueing effect under the horizontal arrangement condition distributed over bilateral parietal-occipital areas and peaked around the right parietal-occipital area, the P1 cueing effect under the vertical arrangement condition distributed laterally over the left parietal-occipital area. Besides, we also observed a central-parietal distributed N1 cueing effect which is slightly different from previous ERP studies of IOR. In short, the results suggest that 1) the inconsistent polarity of the Nd250 in the past ERPs studies is more likely caused by other factors rather than the arrangement of the stimuli. 2) The arrangement of the stimuli could modulate the topographic distribution of the P1 cueing effect rather than the N1 cueing effect.