Abstract Bartlett (1932) using qualitative research methods have demonstrated that memory is a process of construction, and the content of the memory become more reasonable with time. His research is considered to be the beginning of the false memory. DRM paradigm (Deese,1959; Roediger&McDermott,1995) ,a qualitantive research methods, is a classic research paradigm which is widely uesd to induce false memories under laboratory condition. Nonetheless,most of these behavioral memory studies are limited in making such distinctions the component memory processes driving interval-associated memory failures and successes. For example, McDermott (1996) reporter 48h interval between study and test procduced probabilities of false memory that exceeded those of true memory. Alternative neural and cognitive mechanisms may underlie this performance, which could be indexed with functional techniques. The current study investigated the cognitive and neural electrophysiological processes underlying the retrieval phase after imme- diately and 48h delay by means of ERP techniques to further explored the electrophysiological processes of the retention interval effect on false memory. The experiment was conducted with 20 healthy right-handed undergraduates as the subjects and five story contexts as the material during study phase. There were three types of target words during the test phase including higher associated words, critical words and lower associated words. The higher associated words were choose form story contexts; Critical words were new and highly associated with the story contexts, in addition lower associated words were lowly associated with the contexts. The participants studied 5 blocks (one story context in each block and 60 associated target words in each list) and did an old/new recognition test both after short and long delay. Recognition judgment was divided into high confidence level judgment and low confidence level judgment. The results of high confidence level judgment was adopted for statistical analysis. Before long delay test, participants must recall the story context as much as possible for 120s according to story cue. Four subjects were rejected because of the lack of the number of artifact-free trials. The instrument for the statistical analysis used in the experiment is SPSS18.0. The results showed that (1) There was no significant difference on behavior predictions of false recognition between two intervals. False recognition demanded more RT than true recognition on both two intervals. (2) The spatial and temporal distributions of LPC associated with false memory was more positive after long delay relative to short delay.(3) The FN400 elicited by Correct-Rejection recognition was more negative than false memory after both delays. However, both LPC and LSW which were elicited by false memory were more positive than Correct-Rejection recognition.(4) Attenuation of the LPC occurred for false memory relative to true memory after short delay, but there were not different after long delay. （5）The amplitude of LSW elicited by false recognition was greater than true recognition and Correct-Rejection recognition after both delays. In conclusion, the results demonstrated that false memory had showed the different informat- ion processing models between short and long delay. Meanwhile, the different Processing stages between false memory and true recognition were influenced by the time delay. The diversities were recollection-based process and post-retrieval process after short delay, but the diversity only focused on the post-retrieval process after long delay. Key words False Memory, Retention Interval, Story Contexts, LPC, LSW