Information technology is now part of almost every product and service, and the ever-increasing rate of technology and software development demands an appropriate software development process model. The model, whichdescribes the software development process, plays a crucial role for success in software projects. Models are constantly produced and refined in academia and practice, but models from academia tend to remain unused in practice, or take a long time to reach practice. Currently there is a lack of knowledge on how models are transferred or diffused in and between academia and practice, which is the knowledge gap that this study aims to investigate. The research purpose has thus been to investigate and ground a theory about the diffusion of software process models in and between academia and practice. The study is based on data collected from 31 informants, from both academia and practice, and analyzed with a grounded theory method approach. The Diffusion of Innovation theory was consulted during theory development, and the perceived model characteristics, complexity, ando bservability are seen as distinguishing differences between the theories. The resulting grounded theory, “Establishment Patterns of Model Diffusion” describes four important areas, and how they affect each other ina model diffusion situation. The areas are: Motivation; Establishment; Usage; and Quality. How model diffusion takes place depends on the are as themselves, on the prevailing conditions, on the knowledge at hand and on how the areas are evaluated. This shows multi-faceted, time-dependent, and unpredictable patterns for model diffusion. The patterns include experience of model use, the past and future context of use, the characteristics of the model, requirements in the working process, individuals’ empowerment, availability of multiple models, composition of project groups, and the group members’ private and business networks. The findings have several implications. For academia, the benefits may be in the awareness of the multiple patterns for model diffusion and software process improvements, which cannot be separated from each other. Models are constantly adopted and simultaneously modified in practice. Practice may benefit from an awareness of the multiple patterns for model diffusionand sources of experience and knowledge, and how to make use of the existing knowledge capital strategically. Both academia and practice may also benefit from increased cooperation, each contributing their unique experience, and consequently increasing relevance for practice in the diffusion of models, and in developing and receiving research results in a useful format.