To clarify the effects of enforced mental alertness upon spontaneous nystagmus (SPN), the ENG curves of 105 patients were studied. During examination for SPN the recordings were done under three test conditions: I: Patients instructed to "just relax" II: Mental activity alerted by silent mental arithmetics III. Recording after finishing the mental tasks Using seven indices of SPN intensity, mental arithmetics caused a significant improvement in six. The Total Amplitude proved to be the most sensible parameter; there was still a significant improvement even after solving the tasks (test cond. III). A significantly higher rate of improvement has been found in patients with a faint SPN while relaxed, i.e. there was a negative correlation between intensity of SPN and the rate of activation. Patients together with directional prepoderance (DP) had a significantly more intensive SPN than patients without DP or a slight DP only, while there was no difference in the rates of activation in both groups. Central or peripheral lesions did not cause a difference in SPN activation rates. The results give further evidence to the considerable influence of mental alertenes upon SPN; the assumption of a fictitious absolute limit between physiologic and pathologic SPN seems questionable. Mental arithmetics proved to be appropriate to maintain vigilance at a high level.