In the course of experiments on the cerebral evoked potential, recorded by a summation technique through the scalp, it was possible to measure directly the latency of the potential due to afferent conduction time. In two subjects the response following electrical stimulation of the toes was found to occur approximately 20 millisec. later than that following stimulation of the fingers. Using similar stimuli to finger and toe, parallel experiments have been carried out to determine whether there is any corresponding error in the judgement of when two stimuli are simultaneous. The same subjects were presented with pairs of stimuli at a series of intervals up to 100 millisec. and made judgements as to whether they occurred at the same time, or if not, which was first. A control situation, in which the stimuli were given to the two index fingers, was compared with the experimental situation, in which stimuli were presented to finger and toe on opposite sides. The results showed a significant difference in the mean interval at which the two stimuli were judged simultaneous when the conduction pathways were of unequal length, and this difference was in the direction expected should the judgement of simultaneity depend on the simultaneous arrival of the afferent volleys centrally. However, in only one subject was this difference of the order of 20 millisec, as might be expected were the relevant event the onset of the cortical evoked potential. Further, in the other subject the point of apparent simultaneity in the control situation, in which stimuli were given to the right and left index fingers, does not correspond with the point of true simultaneity, as would be expected with conduction pathways of equal length. It is concluded, therefore, that while the temporal relationship of two stimuli is judged on the time of their central arrival rather than on the time of their occurrence at the periphery, judgements may be affected by factors other than conduction delays.