Acquisition of language skills is necessary for social interactions in humans, and speech repetition plays a fundamental role in its acquisition by mapping auditory speech input onto matching speech output. At the level of neuronal networks in the brain, speech repetition occurs within the auditory dorsal stream, which is constituted by a temporo-parieto-frontal network formed by the posterior part of superior temporal sulcus (pSTS), the posterior inferior frontal gyrus (pIFG), and the temporo-parietal junction (Tpj). However, it is less known which component of the dorsal stream plays an essential role in the modulation of speech repetition. To investigate the issue of causality of the dorsal stream in sensorimotor integration of speech repetition, we employed continuous theta-burst transcranial magnetic stimulation (cTBS) in order to disrupt neuronal activity in the stimulated areas. Nineteen right-handed German volunteers performed four behavioral speech tasks. (1) Word-picture matching test: subjects listened to a German noun and were asked to choose a picture which fit to the word heard from four different pictures; the target picture, pictures implying a phonemic or semantic error, and an unrelated picture. (2) Syllable repetition test: subjects listened to one of six different syllables and repeated the perceived syllable immediately. (3) Pseudo-word repetition test: subjects listened to meaningless pseudo-words and immediately repeated it. (4) Sentence repetition test: subjects listened to German sentences and immediately repeated it. In all the tests, error rates (ER) and reaction time (RT) were calculated. Inhibitory cTBS was applied over the individual activated regions of the left pIFG, Tpj, pSTS by using an fMRI-guided TMS neuronavigation system. Left middle occipital gyrus (MOG) was also stimulated as a control region to clarify the topographical specificity of cTBS effects because it is assumed that the MOG is concerned with visual processing. The behavioral speech tests were performed after cTBS over four sites mentioned above, and they were compared by repeated measures ANOVAs. (1) In word-picture matching test, cTBS of Tpj, pSTS and pIFG increased phonemic errors compared to cTBS of MOG. No other types of errors were affected by any sites of cTBS. (2) The syllable repetition ERs were significantly larger after cTBS of Tpj, pSTS and pIFG than after cTBS over MOG. (3) Pseudo-word repetition test demonstrated that ERs significantly increased after cTBS of Tpj and pSTS than after cTBS of MOG and pIFG. (4) In the sentence repetition test, no differences were observed in ERs between sites of cTBS. Throughout all the behavioral tests, RTs were not significantly different between sites of cTBS. Disruption of the dorsal stream led to increases of phonemic errors in word-picture matching test and to increases of errors of syllable and pseudo-word repetitions, demonstrating that the dorsal stream plays a crucial role in sensorimotor integration of phonological processing. The lack of cTBS effects on native sentences may be explained by the idea that the lexical/semantic material processing is done by the ventral auditory stream. The sensory phonological system performs critical modulation of subsequent articulation in the motor system via a sensorimotor interface, supporting evidence that the sensory system is situated hierarchically at higher level to modulate the motor articulation.