Neurophysiological and neuroimaging work has uncovered modulatory influence of long-range lateral connections from outside of the classical receptive field on neuronal and behavioral responses to localized targets. We report two psychophysical experiments investigating visual detection of real and apparent motion in central vision with and without remote and immediate stationary references. At a particular temporal frequency (0.1-12.8 Hz), participants adjusted the amplitude of either triangle-wave (real) or square-wave (stroboscopic/apparent) oscillatory motion of a vertical bar along a straight, horizontal trajectory for the first impression of the target's stationarity/nonstationarity (the displacement threshold). In the relative motion conditions, a stationary reference bar was positioned 23' apart from the target; in the absolute motion conditions, the bar was absent. The thresholds were measured with a dimly-lit uniform background (13 x 13 degrees ) and either in the darkness (experiment 1) or moving-background conditions (experiment 2). For both real and apparent motion, varying the observation conditions yields three sensitivity levels: irrespective of the background, the lowest thresholds occur in the presence of an immediate reference, followed by the moderately increased thresholds obtained with a dimly-lit background alone. The equally high thresholds occur in the darkness and moving-background conditions without any visible stationary references. The results suggest that the spatial frames of reference for visual motion detection are hierarchically nested, yet independent. The findings provide support for the view that absolute motion perception should be considered relative, extending neurophysiological evidence for the existence of long-range lateral connections across the visual field.