Using the DRM paradigm, previous studies found semantical false memory. In recent years, experiments with Chinese characters as materials have been conducted to probe the phonological false memories. However, they haven’t arrived at the unified conclusion. So the present study aims to have a further exploration on the effect of false memory induced by Chinese two-character words’ phonological association. 　　The experiment was carried out on computers using the E-prime software and collectively conducted with 48 undergraduate students as the subjects. In the experiment, subjects were presented with 8 lists of Chinese two-character words and the task was to judge whether each testing item was “old” or not. The E-prime software recorded the results of judgment automatically. There were two types of phonological associates. Half of the lists contain words which belong to first-word associates with the first characters having the same syllables and the last characters having the same initials. While the other half contain words which belong to last-word associates with the first characters having the same initials and the last characters having the same syllables. Test materials included three types of words: studied items which were presented at the study phase; un-presented critical lures which were phonologically associated with the studied items and the intruding items which had no relationship with the studied items. 　　A 2(types of phonological associates) ×3(types of test words) within subjects ANOVA reveal significant main effects of both the two independent variables. For the types of phonological associates , F(1,47)=7.784, p<0.05, and for the types of test words, F(2,94)=474.932, p<0.05. The false recognition rate of the critical lures is significantly higher than that of the non-studied intruding items, and is close to that of the Chinese single-character words, but is significantly lower than the studied ones’. It shows that: (1) Chinese two-character words’ phonological association does have an effect on false memory. (2) The phonological false memory effect induced by the two-character and single-character words are close but are low compared with the semantical and orthographical false memory. (3) False recognition rate for the first-word associates is significantly lower than that of the last-word associates. 　　These results demonstrate that Chinese phonological association can induce false memory, which confirm the findings of Chinese single-character words’ phonological association(Qu & Ding, 2010). But because the activation mechanisms for the semantical and phonological false memory are different. Phonological false memory resulted from low level perceptual-based processing is not so robust and will not be enhanced with the increase of the strength of phonological associates. The present study also suggests that first-word associates and last-word associates have different effects. Future researches need to be done to study the distinctive characteristics on Chinese phonological false memories.