The initial interpretations of motion phenomena are often in terms of eye movements. It reflects a desire to link an objective expression of motion (of the eyes) with a subjective experience of motion. This approach is examined historically with respect to induced motion, motion aftereffects, visual vertigo, autokinetic effects and stroboscopic motion. All but one of these phenomena were initially observed in the natural environment and they have subsequently been studied under laboratory conditions. Eye movements are not correlated with induced motion, the motion aftereffect, or stroboscopic motion, but they do correspond to the direction and extent of visual vertigo. The extent of apparent visual motion in the autokinetic effect far exceeds that of involuntary eye movements but there might be a weak relation between them. If eye movements are to provide a reasonable account of motion illusions then they should apply to all the contours that are visible. This does apply to visual vertigo, but not to induced motion, the motion aftereffect and stroboscopic motion, all of which involve relative motion in a visible display. The autokinetic phenomenon involves the perception of isolated stimuli, and so interpretations in terms of some internal eye movement signals remain sustainable.