In our previous study, continuous visual (cartoon and random dot motion) and auditory (music) stimulation changed the somatosensory evoked magnetic fields following electrical stimulation of the median nerve in human subjects. They enhanced the middle-latency components (3M and 4M) generated in the contralateral primary somatosensory cortex, and reduced the MI component generated in the ipsilateral secondary somatosensory cortex. We speculated that such interference effects were caused by activation of polymodal neurons in areas 5 and/or 7 of the parietal lobe as well as in the medial superior temporal region and superior temporal sulcus. However, we could not exclude the effect of attention on such interference effects. In the present study, to know the effect of attention on visual and auditory interference in these changes, we stimulated the bilateral median nerves unilaterally in a random order, and asked subjects to count the number of times the left median nerve was stimulated while visual or auditory interference was applied. Five components (1M-5M) were identified in the hemisphere contralateral to the stimulated nerve and only one component (MI) was found in the ipsilateral hemisphere. The 3M and 4M components (33-75 ms in latency) were enhanced by both attention and visual interference stimulation but not by auditory stimulation. The 5M component (70-115 ms) and MI component (70-133 ms) were enhanced by attention, but were not changed by attention together with visual or auditory interference. Summarizing the results of our previous study and the present study, (1) visual interference alone enhanced the 3M and 4M but reduced the MI, and (2) attention alone also enhanced the 3M and 4M, and enhanced the 5M and MI. As a result, (3) visual interference with attention enhanced the 3M and 4M more, and showed no significant change of the 5M and MI. This was compatible with a summation of the effects caused by visual interference alone and attention alone, but some interactions between visual interference and attention might have taken place. The locations of dipoles of all components were not significantly changed by attention or any interference stimulation. These findings support the idea that there are significant interactions of activities relating to somatosensory stimulation, visual stimulation and cognitive function, in both the primary and the secondary somatosensory cortex in humans.