This masters thesis examines an important field, the field of software development, in the technology driven world of today where computers running software systems have achieved a critical role. The everyday lives of people are affected, when software is developed that doesn’t fit its intended purpose. I posit that the fitness of software is very much affected by the way it is developed, and that improvements in this fitness won’t come from better technology, but can be achieved when the software development process is improved. Specifically, this thesis examines the values and basic principles of the development methodology known as Extreme Programming, one of the new agile development methodologies, from a learning perspective by viewing the developers as members of a Community of Practice (CoP), as described by Etienne Wenger. In the thesis, I attempt to determine how the perspective can contribute to our understanding of the software development process, when Extreme Programming (XP) is implemented, and if it can show aspects of the XP methodology that can be improved to facilitate improvement in the learning process. I base my conclusion upon reflections on XP as a development methodology, and Communities of Practice as a description of how people learn in practice, and data collected by questioning experienced software developers, who use Extreme Programming to guide their development process. I come to the conclusion that software developers using XP can viewed as a community of Practice, there is an excellent mapping between the values of XP and CoP, and that the CoP viewpoint does point towards a number of areas in the XP methodology that can probably be improved, by better facilitating the learning process, and hopefully thereby improving the fitness of software developed using the methodology.