In this study, we examined whether visual context can be learned through a dynamic display and whether it can facilitate sustained attentional tracking by combining a multiple object tracking (MOT) task and a contextual cueing procedure. The trajectories of the targets and distractors in the MOT task were made invariant by repeatedly presenting them. The results revealed that when the targets were repeatedly displayed, tracking performance implicitly improved, and this effect was enhanced when the unattended distractors in the displays were also repeated. However, the repetition of the distractors alone did not produce any effect. Interestingly, when the targets and distractors were switched in a display in which the distractors had been previously repeated, the tracking performance was impaired as compared with that in the case of nonrepeated displays. We concluded that the contextual information in a dynamic display facilitates attentional tracking and that different types of contextual modulations occurred in MOT processes, such as facilitation for attended targets and inhibition for ignored distractors.