Analysis of spontaneous nystagmus is important in the evaluation of dizzy patients. The aim was to measure how different visual conditions affect the properties of nystagmus using three-dimensional video-oculography (VOG). We compared prevalence, frequency and slow phase velocity (SPV) of the spontaneous nystagmus with gaze fixation allowed, with Frenzel's glasses, and in total darkness. Twenty-five patients (35 measurements) with the peripheral vestibular pathologies were included. The prevalence of nystagmus with the gaze fixation was 40%, and it increased significantly to 66% with Frenzel's glasses and regular room lights on (p < 0.01). The prevalence increased significantly to 83% when the regular room lights were switched off (p = 0.014), and further to 100% in total darkness (p = 0.025). The mean SPV of nystagmus with visual fixation allowed was 1.0°/s. It increased to 2.4°/s with Frenzel's glasses and room lights on, and additionally to 3.1°/s, when the regular room lights were switched off. The mean SPV in total darkness was 6.9°/s. The difference was highly significant between all test conditions (p < 0.01). The frequency of nystagmus was 0.7 beats/s with gaze fixation, 0.8 beats/s in both the test conditions with Frenzel's glasses on, and 1.2 beats/s in total darkness. The frequency in total darkness was significantly higher (p < 0.05) than with Frenzel's glasses, and more so than with visual fixation (p = 0.003). The VOG in total darkness is superior in detecting nystagmus, since Frenzel's glasses allow visual suppression to happen, and this effect is reinforced with gaze fixation allowed. Strict control of visual surroundings is essential in interpreting peripheral nystagmus.