Purpose: A head tilt towards the shoulder (roll) induces an ocular counter-roll (OCR), i.e. torsion in the opposite direction to the head. How this counter-rolled position is maintained during a static head tilt is in debate. In a previous study, we reported an OCR-increasing drift subsequent to the head tilt. This finding is in contrast to other reports where no such response was found. The primary aim of this study was to repeat the experiment during a prolonged head-tilt test and to describe the OCR characteristics. A secondary aim was to investigate the influence of spatial visual cues on OCR. Methods: Five male subjects performed a head tilt (30°) towards the right shoulder while the eye position was recorded during a 10-minute interval. In test 1, the subjects viewed a target with no cues for spatial orientation. The same head-tilt paradigm was repeated in test 2 with a visual target with spatial cues. Two samples of data were extracted from the start and the end of the recordings for statistical analysis. Results: Subsequent to the head tilt, a slow OCR-increasing drift in the opposite direction to the head roll was found in all subjects. On average, this drift lasted for 30 sec (± 5) in test 1 and for 55 sec (± 18) in test 2. The drift was then found to change its direction, i.e. the eyes were rotated in the same direction as the head roll. When measured after 10 minutes, the OCR was significantly decreased. Conclusions: The OCR during static head tilt is not constant. During the first minute there is a gradually increasing OCR. Thereafter, the amplitude of the OCR decreases gradually. These changes are influenced to some extent by spatial visual cues. Possible mechanisms are adaptive responses in otolithic afferents as well as central nervous memory functions related to the semicircular canal system.