The dynamic and haphazard nature of real-world visual experience requires that object representations be maintained through extrinsic interruptions, such as when attention is allocated to other objects and events. What resources underlie and limit this ability? These studies demonstrated that even infants’ persisting object representations can be maintained over brief interruptions from additional independent events, just as one’s memory of a traffic scene may be maintained through a brief glance in the rearview mirror. However, this ability is nevertheless subject to an object-based limit: if the same interrupting object is simply segmented into 4 (or even 2) objects, then it will impair the maintenance of other persisting objects.