ABSTRACTWaveform patterns evoked by 4 intensities of flash in normal subjects were studied in relation to intersubject variability. Time‐frequency distribution curves of all peaks occurring between 11 and 280 msec after flash onset and meeting minimal criteria were obtained from 46 males. These distributions closely corresponded to similar data reported by others for single intensity stimulation. An algorithm was developed which identified in 67 to 100% of instances a single “peak event’ within the time ranges of each of 6 peak distributions. Many peak events appeared and disappeared within the 4 intensity sets of individuals. Latencies were obtained for these peak events. Application of the algorithm to a replicate sample of 29Ss, which included 8 females, indicated generalizability. Test‐retest data on 15Ss showed its reliability. The data suggest that methodology significantly contributes to the variability of peak identification among subjects. This may be reduced by employing multiple intensities of stim