A series of experiments test the recent claim that the hippocampus is necessary for the binding of features in working memory. Some potential limitations of studies underlying this claim are discussed, and an attempt is made to further test the hypothesis by studying a case of developmental amnesia whose extensively investigated pathology appears to be principally limited to the hippocampus, and who shows the expected deficit in episodic long-term memory. One series of experiments studied the short-term visual binding of color and shape under conditions ranging from simple presentation of colored objects through the more demanding task of combining the features when separated in space, or in time. A second set of experiments studied the capacity to use sentence structure to bind words into chunks in short-term verbal memory. Hippocampal pathology did not lead to a decrement on any of these tasks, suggesting that the hippocampus is not essential for short-term binding in working memory. (Contains 2 figures.)