The VEPs to checkerboard-flashes were recorded from Cz in 32 healthy adult volunteers (14 men, 18 women) under biofeedback conditions. The subjects were instructed to modify N100-P200 component in such a way that the amplitude of the segment within selected time-window from the beginning of the sample to its main positive peak exceeded the definite threshold. Each subject solved the task by means of individual combination of voltage and temporal regulation with the prevalence of the latter. Shifts of the latencies both of N100 and P200 were involved in VEP modifications. Four groups of subjects were distinguished by their success in task performance which depended on the strategy and level of self-estimated tiredness. The latency of N300 (out of the chosen time-sample) was longer for the subjects which were successful in increasing the number of "correct" responses. Differences in VEP components depended on the strategy types, levels of self-estimated tiredness, and sex on the subjects.