The binding problem has been discussed mainly in visual perception. If the binding in perception is carried out sequentially by selective attention, however, visual short-term memory (VSTM) faces the binding problem to recognize multiple objects simultaneously. Despite some theoretical accounts and empirical data for maintenance of multiple feature-bound objects, recent studies directly testing the binding in VSTM suggest that the capacity of feature-bound objects may be highly limited, consistent with an idea that binding is sequential even in VSTM. Binding in perception and VSTM may share a common mechanism operating on feature maps at lower- and higher-level visual areas, respectively.