When knowledge management (KM) began to emerge in the late 1980s, it was seen as an innovative solution to the problems of managing knowledge in a competitive and increasingly internationalized business environment. At that time, the term was often used in conjunction with so-called expert systems that dealt with hard1, structured knowledge (Hildreth, Wright & Kimble, 1999). During this period, knowledge was seen as something that had an independent existence; it could be captured from an expert, codified in a series of rules, and stored in a computer. However, many authors have argued that, in practice, KM was often little more than information management systems rebadged (Wilson, 2002). More recently, there has begun to be recognition of the importance of softer, less structured types of knowledge (Hildreth, Wright & Kimble, 1999). There has been a growing awareness that knowledge is not found in rules, frames, cases, predicate logic, or document repositories but that other factors were at work. This inevitably raises questions about what these other factors are and how this new softer form of knowledge might be managed. Communities of practice (CoPs) were identified by many as a means by which this softer type of knowledge could be created, shared, and sustained. From this, it was a small step to arguing that CoPs were in fact a new approach to KM that offered the solution to many of the shortcomings of the earlier, systems based attempts at KM. However, the concept of a CoP is built around a very different set of principles to those put forward by the proponents of KM, and it is not always clear that this argument will hold. Much of what is now called KM has developed in a formal organization setting. In this setting, groups are often seen simply as collections of people who are brought together to complete a specific task; once the task has been completed, the group can be dissolved. These groups are often created in a top down fashion, and the structure of the group usually reflects the existing organizational hierarchy. The successful completion of the task (or repeated series of tasks) is usually the basis for financial or other reward. In contrast, CoPs tend to be self-perpetuating and self-directed. The focus of a CoP is not on a narrowly bounded task but on a living and dynamic practice; the rewards are intrinsic rather than financial. Authority and legitimacy are not a function of formal rank or hierarchy but of an informal status in the group. In summary, the members of a CoP have more in common with a troop of altruistic volunteers than a band of paid employees. This contrast between the nature of CoPs and the demands of a high tech, global commercial enterprise raises two important questions that we will return to in the Communities of Practice Today section. First, do CoPs really offer a way to manage the softer aspects of knowledge? That is to say, can they be initiated and directed by management, or will the outcome always be the product of the emergent properties of a self-directed and self-organized group? Following on from this, the second question is: if they do offer ways to manage the softer aspects of knowledge, will they work in today’s high tech and increasingly internationalized virtual world?