[en] Objective: To localize the language cortex associated with Chinese word processing by magnetic source imaging (MSI). Methods: Eight right handed and one left handed healthy native Chinese speakers, including 5 men and 4 women, aged from 14 to 32 years, were examined by magnetoencephalography (MEG) and 1.5 T MR unit. All subjects were given 50 times pure tone stimuli (intensity was 80 dB sound pressure level), then 150 pairs of Chinese words (the meaning of the words was related or not related) auditory stimuli (intensity was 80 dB sound pressure level), and then 50 times pure tone stimuli at last (intensity was 80 dB sound pressure level). Evoked response fields (ERFs) time locked to the pure tone and Chinese words were recorded in a magnetically shielded room using a whole-head neuromagnetometer (Model Vectorview 306, made by 4-D Neuroimaging company, Finland) in real-time. The acquired data were averaged by the acquisition computer according to the response to the pure tone, related pairs of words and not related pairs of words. The data obtained by the MEG could be superimposed on MRI. Results: There were two obvious higher magnetic waves named M50 and M100 (two peaks occurred about 50 ms and 100 ms after giving the subjects binaurally stimuli). M50 and M100 in all subjects were localized in the bilateral transverse temporal gyri. The responses to the pairs of Chinese words (the meaning of the words was related or not related) were similar in the same hemisphere of the same subjects. There was a higher peak during 300-600 ms in the right hemisphere in the left handed subject, but there was no peak during 300-600 ms in his left hemisphere. It indicated that the language dominant hemisphere localized in the right hemisphere. Superimposing the MEG data on MRI, the language area was localized in the Wernicke's areas. There were two 300-600 ms response peaks in the bilateral hemispheres (the amplitude of the 300-600 ms response peaks in the bilateral hemisphere was almost the same) in two right handed subjects. It implied that the language area localized in the two hemispheres in those two subjects. There was one peak in each hemisphere (300-600 ms response) in 6 subjects, but the amplitude of the wave in left hemisphere in the 6 subjects was much higher than that of the right hemisphere. After giving the auditory task in all subjects and choosing several time points from the later component (300-600 ms response) randomly and superimposing on MRI, all these time points localized in the posterior part of the superior temporal gyri, that was Wernicke's areas. Conclusion: Comparing the later components of the bilateral hemispheres, the amplitude of wave in the language dominant hemisphere is much higher than that of the non-language dominant hemisphere. The judgement of the Chinese words (the meaning is related or not related) could be used in identifying the language areas