To determine if linguistic processing could be selectively disrupted with repetitive transcranial magnetic stimulation (rTMS), rTMS was performed during a picture-word verification task. Seven right-handed subjects were trained in two conditions: picture-word verification, which required the subject to verify whether the picture of an object matched the subtitle name on the same page, and frame verification, which required subjects to verify whether there was a rectangular frame around the combined object picture and subtitle. Half of the trials were performed during rTMS. The effects of rTMS on performance were evaluated at the following four scalp positions: left anterior (the area where rTMS produced speech arrest), a mirror site on the right, and two positions in the left and right parietal region. Stimulation over the left deltoid muscle served as a control. Subjects had less difficulty in making picture-word matching decisions during unstimulated compared with stimulated trials at the left anterior and posterior positions. No significant difference in accuracy was detected in the frame verification condition, but response times in the frame verification condition were longer with stimulation at the left anterior position. Because rTMS of the dominant hemisphere affected linguistic processing independent of speech motor output, we confirm that rTMS may be used to investigate language and other cognitive functions.