This chapter examines nonverbal and verbal literature; After evidence is presented and some theoretical issues are explored, a recent study by Cowan, Chen, and Rouder will be used to examine the hypothesis that there is a fixed capacity limit in terms of chunks recalled. Keeping in mind the many different kinds of limits to immediate memory that have been explored, let us return to the concept of capacity as it relates to serial recall and the work of Miller. Performance after recoding might involve three times the number of digits per chunk, but probably less because it may prevent some of the chunking together of digits that ordinarily might take place without the learned recoding. As useful as Miller's notion of chunking is, it has to be viewed as somewhat of an oversimplification of how knowledge helps working memory. The initial plan for an independent measure of chunking in the results of Cowan.