The allocation of attention during reading has long been a central focus of research. A key question concerns the extent to which lexical processing is influenced solely by the difficulty of currently fixated words, as proposed by the serial attention hypothesis, or whether it is also influenced by the difficulty of upcoming words (i.e., parafoveal-on-foveal effects [PoF effects]), as proposed by the attention gradient hypothesis. The present study provides oculomotor and neural evidence for how mislocated fixation can influence PoF effects in Chinese reading. We manipulated preview words following the target words (repeated vs. original) and categorized the two-character target words into two fixation position groups (Character 1: no fixation error group vs. Character 2: potential mislocated fixations group). Experiment 1 used an eye-tracking technique, and Experiment 2 coregistered electroencephalogram and eye tracking with a larger character size to validate the findings. Overall, the results indicated that inhibitory repeated PoF effects were observed only at Character 2, as evidenced by fixation durations and fixation-related potentials. These findings provide novel eye-movement and neural evidence suggesting that inhibitory repeated PoF effects may, at least in part, be related to mislocated fixations caused by fixations on posttarget words that are mistakenly recorded as being at the end of target words due to measurement (or saccade) errors, while attention remains focused on repeated preview. Taken together, these findings suggest that the mislocated fixation account, situated within the serial attention hypothesis, provides a viable explanation for interpreting the PoF effect without requiring parallel processing. (PsycInfo Database Record (c) 2025 APA, all rights reserved).