This Ph.D. thesis analyses open source software from an economical standpoint and answers the meta-question: "Why is open source software being developed?" The question is motivated by the apparent success of open source software in the market place and the fact that both individuals and firms use and develop open source software. The thesis presents a review of some of the existing litterature and konsludes that this litterature is insufficient to account for the bahaviour of firms. The thesis then presents three theoretical perspective, which are to serve tools for answering the meta-question. A model for open source software development is developed, which focuses at the leavel of the individual software project. The model is derived from the practical development process coupled with economics using the theoretical tools. The model is further coupled with three different software licenses the GPL, BSD and Microsoft EULA. The model is used to explain agents behavior given the different licenses. The model is then tested using a number of mini cases describing open source software development stories. These stories are produced from a number of indepth interviews with open source software developes. The test concludes that the model does in fact help to explain agents bahaviour.