Different theoretical views exist regarding whether false memories contain perceptual information or are merely conceptual in nature. To address this question, we conducted three experiments to examine whether false memories for pictures had a priming effect on a perceptual closure task. In Experiment 1, participants were presented with pictorial versions of Deese/Roediger-McDermott (DRM) lists and received a recognition task. Finally, in the perceptual closure task (PCT), participants were shown degraded pictures (studied pictures, critical pictures, unrelated pictures) that became clearer over time and had to identify the object depicted as quickly as possible. The results showed that false memories for pictures did not exhibit a priming effect in the PCT. Specifically, picture identifications based on false memories for visual stimuli were significantly slower than those based on true memories and the former did not differ from that of unrelated items. In Experiments 2 and 3, we manipulated the modality (verbal vs. pictorial) of the study phase and the PCT phase. In both experiments, false memories for pictures primed pictures significantly slower than true memories in the pictorial PCT, but false memories for pictures primed words faster than true memories in the verbal PCT. Our results suggest that false memories for pictures are unlikely to contain perceptual information but rather that they are conceptual in nature.