Objectives: To determine whether mental alerting (MA) makes a significant difference to nystagmus recorded during gaze testing using electronystagmography (ENG). Design: A prospective study, blinded interpretation of test results, alternate order of testing sequence. Setting: Vestibular assessment clinic in a tertiary referral centre. Participants: A total of 80 patients were seen by the authors (R.L.H., K.F.L.M.) for vestibular assessment, between August 2002 and February 2004. Patients cognitively unable to perform MA tasks or with limited English were omitted. Patients all underwent gaze-testing using ENG. Each patient was tested twice, once without MA and once with an MA task, in an alternate test order. Main outcome measures: Slow-phase velocity values of recorded nystagmus, number of patients with clinically significant nystagmus (slow-phase velocity greater than or equal to6 deg/s). Results: A total of eight patients had significant gaze-evoked nystagmus when tested without MA and a total of six patients when tested with MA. Only four of these patients were common to both groups. When all nystagmus values (regardless of significance) were considered, no statistically significant differences between nystagmus measured with and without alerting were found (Wilcoxon signed rank test, P > 0.05). Conclusions: Mental alerting during gaze testing did not result in a significant difference in nystagmus and is just as likely to result in an exacerbation as a reduction in slow-phase velocity. In addition, MA is likely to result in a noisy (and therefore difficult to interpret) recording and increases the impact of the test on the patient. In view of these findings, we propose that MA is contraindicated during the gaze-testing component of the vestibular test battery.