Intrinsic signal imaging from inferotemporal (IT) cortex, a visual area essential for object perception and recognition, revealed that visually presented objects activated patches in a distributed manner. When visual features of these objects were partially removed, the simplified stimuli activated only a subset of the patches elicited by the originals. This result, in conjunction with extracellular recording, suggests that an object is represented by a combination of cortical columns, each of which represents a visual feature (feature column). Simplification of an object occasionally caused the appearance of columns that were not active when viewing the more complex form. Thus, not all the columns related to a particular feature were necessarily activated by the original objects. Taken together, these results suggest that objects may be represented not only by simply combining feature columns but also by using a variety of combinations of active and inactive columns for individual features.