Using a reading comprehension task, we explored whether body postures would influence mind wandering, a universal internally self-generated activity. Specifically, participants were instructed to perform a reading comprehension task under three postural conditions (lying supine, sitting, and standing upright). Probe-caught technique with prompts presented at irregular intervals was adapted to measure the frequency of mind wandering. Self-caught method was used to measure the meta-awareness of mind wandering by self-reports. Results indicated that the radio of mind wandering was significantly greater in lying than standing and sitting, but the meta-awareness of it was not different among three postures. Moreover, the reading performance, an indirect indicator of executive control, decreased in lying compared to standing and sitting. We suggested that the increase of mind wandering in lying posture may due to the dysfunction of executive control, which also results in the redistribution of cognitive resources. Suggestions for future research are proposed.