When attempting to memorize action sentences (e.g., open an umbrella), performing the action of the sentence (enacted encoding) results in better memory performance than simply memorizing the sentences (verbal encoding). This memory enhancement is called the enactment effect. Magnetoencephalography (MEG) was used to elucidate whether the enactment effect is due to physical motor information or whether movement representation is the critical factor in the enactment effect. Physical motor information, which is implicated in the primary motor cortex, represents the speed, form, and kinematic sense of a movement, while movement representation indicates semantic and conceptual information such as movement formulae, movement ideas, and movement imagery, which are especially associated with the parietal cortex. We measured activities within the motor region and parietal cortex during a recognition test and compared activities during recognition with enacted and verbal encoding condition. The results showed that recognition performance was better for enacted encoding. The MEG data indicated that the left primary motor cortex with enacted encoding condition was activated in all subjects, though with verbal encoding condition, this activation appeared in only one subject. These activities were observed between 150 and 250 ms after recognition stimuli onset and were transmitted into the left parietal cortex. Moreover, activities in the right parietal cortex following enacted encoding were greater than those following verbal encoding, and the activities appeared 600-700 ms after onset of the recognition stimuli. These results suggest that the enactment effect occurs by the reactivation of the physical motor information and that this information facilitates activities related to movement representation.