The present study examined the influence of word complexity on the divided-word effect. By manipulating presentation conditions and visual complexity, we found a significant divided-word effect that the reading times such as gaze duration and total reading time were significantly longer in the divided-word presentation condition than in both the line-final and line-initial presentation conditions. On the measure of total reading time, the marginally significant interaction between the divided-word versus line-final presentation comparison and complexity showed that the divided-word effect was (...) larger for low complexity words than that for high complexity words. These results suggest that dividing a word across two lines interferes with reading, and word complexity modulates this effect. (shrink)