An extensive literature has demonstrated that encoding information in a self-referential manner enhances subsequent memory performance. This self-reference effect is generally elicited in multiple paradigms. Klein and Loftus (1988) developed a dual-processing explanation of self-referent encoding. They emphasized that both elaborative and organizational processes are involved in the self-referential facilitation in true memories. However, a number of prior research has demonstrated that organizational processes could induce high level of false memories. With two experiments, the present study used Deese-Roediger-McDermott paradigm to examine the potential influence of self-reference on false recognition. Across two experiments, participants were instructed to remember lists of words and complete a following recognition test immediately. In study phase, list words were presented to participants one by one in the center of the screen at the rate of one word per 1.5 s paired with different cues. In experiment 1, participants viewed DRM lists in blocked design, with one third of DRM lists paired with participants’ own name (self-referential condition), another one third of DRM lists paired with the name of a well-known individual (other-referential condition), and the last paired with a colored block (neutral condition). Participants were told to remember both the words and cues. In experiment 2, all list words were presented randomly to participants in interleaved design. In test phase, participants were asked to make judgments on whether or not the test items were presented before and a R/K judgment on each test item. The results showed that, when studied in a self-referential manner, the list words were better remembered than in other-referential and neutral conditions. Most importantly, the false recognitions of critical lures were higher in self-referential condition than other-referential and neutral conditions. When list words were presented in a random order, self-referential effect in false memory decreased significantly. In other words, disrupting organizational process reduced false memories cross all conditions, but self-referential effect in false memory did reliably exist. In addition, both for true memory and false memory, recollection in self-referential condition was significantly higher than other-referential and neutral conditions, whereas there were no differences in familiarity across three conditions. The results demonstrate that self-reference could facilitate both true memory and false memory in DRM paradigm. The influence of self-reference on false memory could be interpreted through the promoted organizational process between studied items and critical lures. The dissociation between recollection and familiarity suggests that recollection and familiarity are two independent systems, and recollection might play a more important role in self-reference processing than familiarity.