Decades of work on dual-task performance and selective attention has provided clear and robust evidence that dividing attention (DA) during encoding generally impairs subsequent memory performance. However, recent research reports that detection of infrequent targets in a DA condition while encoding a series of images into memory can enhance memory of images on subsequent memory test. This unusual phenomenon is described as the attentional boost effect (ABE). Previous research has explored the mechanism underling the ABE, and a lot of evidence showed that the ABE primarily enhanced perceptual encoding of the target-paired items, leading to a robust memory advantage for the target-paired items. However, some results argued against the perceptual encoding hypothesis of the ABE (Mulligan et al., 2014). Thus, the present experiment explores whether the memory benefit of the ABE is attributed to the perceptual encoding of the stimulus or to the later phase of elaboration and rehearsal. To investigate the neural activity of ABE, event-related brain potential (ERP) technique was adopted to record electroencephalographic (EEG) data of 24 participants who engaged in a second target-detection task while encoding a long sequence of words. A total of 600 words were presented during the encoding phase. Among these words, 200 words were presented with target indicators (target words), 200 words presented with distractor indicators (target words), and others were presented without any indicators (baseline words). Participants were told to read aloud each word while monitoring a target indicator, and withhold the response when they saw a distractor indicator, or did not see any indicators. Then a recognition task was given to test the participants’ memory of the words. Recording and analyzing the behavioral data and ERP data in the encoding and recognition test phase. Behavioral results showed that words presented in the target trials were more accurately recognized than those presented in the baseline (p<0.001), or distractor trials (p<0.001). There was no significant difference in the memory of words between the distractor and baseline conditions (p=0.12). The ERP results showed that, during the study phase, baseline trials induced a larger negative N1(130-180ms) amplitude, compared with target and distractor trials, and there was a significant difference between target and distractor trials in N2 (250-350ms), P300 (350-500ms), and later 700-1000ms time windows. During the test phase, the ERP old/new effects (FN400) induced by target words and baseline words were mainly manifested in the anterior half of the brain for the 300-500ms time windows. In addition, we also found the difference of ERP between target words and distractor words was embodied in that target words produced a more positive ERP waveform than distractor words and went on to 800ms. In conclusion, the difference of ERP between target words and distractor words is not only reflected in the ERP components related to perceptual processing, but also in the ERP components related to later conscious recollection. These data lead us to conclude that ABE could enhance the perceptual encoding of the stimulus, and the later phase of elaboration and rehearsal.