It has been shown that transcranial magnetic stimulation can delay simple reaction time; this happens when the stimulation is delivered during the reaction time and over the cortical area which commands the prime mover of the required response. Although it is agreed that magnetic stimulation could be a useful tool for studying information processing in man, we argue that, because of the use of simple reaction time, the results reported so far are difficult to interpret within this theoretical framework. In the present paper, three experiments are reported. Six subjects participated in experiment 1 in which magnetic stimulation was delivered, at different times, during choice reaction time. The effects of the magnetic stimulation of the cortical area involved in the response (induced current passing forward over the motor representation of the responding hand), were compared to the effects of an electrical stimulation of the median nerve (K-reflex). In a first control experiment (experiment 2a; 5 subjects), the coil was placed over the ipsilateral motor cortex (induced current passing backward over the motor representation of the non-responding hand) thus minimizing the probability to excite the same neural nets as in the first experiment. In a second control experiment (experiment 2b; 4 subjects), the coil was placed a few centimeters away from the subject's scalp thus ensuring no stimulation of any neural nets. The results show that: (1) the noise and the smarting of the skin associated with the coil discharge produce an intersensory facilitation thereby shortening reaction time (experiment 2a), (2) actually, the noise produced by the stimulation is sufficient to produce such a facilitatory effect (experiment 2b), (3) a stimulation over the area at the origin of the motor command causes a reaction time delay which counteracts this intersensory facilitation effect (experiment 1), (4) in this latter case, the closer the stimulation to the actual overt response, the longer the delay and (5) there is no trace of correlation between the amplitude of the motor evoked potential and the reaction time change. (C) 1997 Elsevier Science B.V.