Multistable perception occurs when a single, but ambiguous stimulus drives perceptual alternations. Understanding its mechanisms has a direct impact on perceptual inference and decision making. A model proposed by Shpiro and colleagues explains the dynamics of bistable perception by neural adaptation and driving noise. The action of adaptation and noise on competing neuronal populations—each encoding a perceptual representation—results in perceptual reversals.Goal. To test effects of noise and adaptation on perceptual reversal speed.Methods. We manipulate noise and adaptation using predictability of the retinal projection of the stimulus. A Necker cube was presented to 16 observers instructed to gaze at a central fixation cross while reporting their perceptual changes by key press. The stimulus followed either a smooth, predictable motion; a pseudo-random motion; or no motion at all (control). Our hypotheses predicted higher (lower) reversal speeds for low (high) predictable motion w.r.t. no motion.Results. Key press analysis validated our hypothesis for unpredictable motion, but not for predictable motion. We explain the latter by quantifying correlations between stimulus and gaze positions. This shows that observers executed micro-pursuit-like movements under predictable stimulus motion, thereby increasing the effect of adaptation on reversal speed w.r.t. our hypothesis.