We used repetitive, rapid-rate transcranial magnetic stimulation (rTMS) for the noninvasive study of visual attention in humans. Six right-handed volunteers completed eight blocks of 20 single- and 10 double-visual-stimulus trials. The visual stimulus was a single asterisk on the right or left side of a computer screen or two asterisks presented simultaneously. The subject had to respond to the stimulus by pressing the right or left response key or both keys simultaneously. During six of the blocks, we applied focal rTMS in trains of five pulses at 25 Hz and 115% of the subject9s motor threshold intensity to scalp positions O1, O2, P3, P4, T5, or T6. Occipital rTMS led to a large number of misses of the contralateral asterisk regardless of whether a single or double stimulus was presented. Parietal rTMS did not induce misses of single stimuli but led to a large number of misses of the contralateral asterisk in the double-stimulus condition. The effects of temporal rTMS were inconsistent. We conclude that rTMS to the occipital lobe causes a sensory detection block, whereas rTMS to the parietal lobe can induce selective extinction of contralateral visual stimuli during a simultaneous double stimulus.