Numerous neurolinguistic works on the correlation between syntactic movements and brain regions have been conducted under same hypothesis that syntactic operations are localizable (Grodzinsky, 2006a). Although various aspects of syntactic movements were concerned, including (Ben-Shachar et al., 2003; Makuuichi et al., 2013; Shetreet & Friedmann, 2014) and the distance between the moved phrase and the trace position has also been discussed (Makuuichi et al., 2013), the direction of syntactic movements (leftward vs. rightward) draws little attention, despite the fact that syntactic dislocations do not always involve leftward movements. We thus conducted an event-related fMRI experiment to explore biological validity of the leftward and rightward syntactic movement in Chinese. The stimuli were designed under the four conditions: (1) topicalized sentences (TOP), which involves the object to move leftward to sentence-initial position, (2) object relative clauses (ORC), which require the object of relative clause to move rightward to the edge position, (3) subject relative clauses (SRC), with the subject of RC moving rightward, and (4) the baseline (BASE), the sentence with the typical SVO word order. There were 166 reading stimuli in total (31 sentences for each condition, with 42 fillers), and the Latin-square was adopted to create 4 lists for counter-balance. Fifty comprehension questions were given to ensure participants’ attention. We collected data from 30 participants (male: 11, female: 19). The statistical analysis of one-way ANOVA (with the p value < 0.01) revealed that the contrast of left/right directions was significant. The TOP > ORC (i.e. leftward > rightward) contrast showed greater activations in the left posterior middle temporal gyrus, left inferior parietal lobule, left supramarginal gyrus, and right visual-motor coordinated area (BA 7). In contrast, the ORC > TOP contrast (i.e. rightward > leftward) involved greater activations in the left lingual gyrus, left anterior and central middle temporal gyrus, left hippocampus, and right premotor cortex (BA6). Furthermore, the results of TOP > SRC / SRC > TOP contrast pair were close to the TOP > ORC / ORC > TOP pair results with slight differences. The result suggested that leftward and rightward dislocations recruited saliently different brain regions. We attributed such differences to the order of gap and filler processing in the brain. In leftward movement, the moved phrase, also known as the filler, is read first and temporally stored until the gap (also known as trace in syntax) is reached. On the contrary, the gap is processed before the filler in the sentence involving rightward movement. However, how the order of gap and filler processing correlates with the regions with greater activations deserves further consideration.