The effects of magnetic coil (MC) stimulation of human visual cortex on the foveal perception of briefly presented letter trigrams include: (1) letters were nearly always reported correctly at visual stimulus-MC pulse intervals less than 60-80 msec or greater than 120-140 msec. Thus, by 120-140 msec, information related to letter recognition is relayed from calcarine cortex. (2) Presentation of equiluminant chromatic stimuli (specifically green letters against a red background) results in suppression curves which commence at longer latencies than those obtained with achromatic stimuli. (3) At a stimulus-MC pulse interval of 100 msec, shifting the MC laterally or rostrally resulted in suppression of the contralateral or caudal-most letter respectively. This implies a focal, topographical effect on visual cortex. (4) Two trigram stimuli separated in time (e.g. 100 msec) resulted in classical backward masking in which S1 (the target) was suppressed by S2 (the mask), using an S2/S1 luminance-contrast ratio of 4:1. When the MC was subsequently discharged 80-100 msec after S2, and S2 was suppressed, the response to S1 was easily retrieved (unmasked). Presumably, by 160 msec, S1 has been transmitted to the next processing, extrastriate level. (5) The unmasking phenomenon has been used to track information flow from visual cortex to higher cortical centers (e.g. Wernicke's, Broca's, and related areas). (6) Using a prototype repetitive stimulator, a consecutive train of single MC pulses given 70, 143 and 216 msec following a brief alphabetic trigram stimulus elicited a significant reduction in letter perception. This notably contrasts with the absence of suppression when a single MC pulse was given 70 or 143 msec following presentation of the alphabetic trigram. The results with 3 pulses suggest that the first MC pulse (at 70 msec) delays but requires repetition to prevent processing and/or transmission of information from visual cortex.