Many companies around the world have implemented Six Sigma as a problem solving methodology especially useful for dealing with recurring problems in business processes. Since the 1980s when it was developed at Motorola, many companies have tried to implement Six Sigma to fit their own company’s culture and goals. This thesis presents a longitudinal case study describing the evolution of ‘Six Sigma Management’ at Siemens in Sweden. The success of the programme was to a large degree built on previous failures, confirming Juran’s old saying ‘Failure is a gold mine’. From the case study, success factors for implementing Six Sigma at Siemens are identified and compared to those given in the literature. Some of the most critical success factors identified at Siemens had not been mentioned as such in the literature before. The main conclusion of the study is that, in order to succeed and get sustainable results from a Six Sigma programme, Six Sigma should be integrated with Process Management, instead of just running Six Sigma as a separate initiative in an organisation. Furthermore, the thesis includes papers presenting methods and tools to be used in a Six Sigma programme or in Six Sigma projects. They deal with: how to identify suitable Six Sigma projects, how to select which Six Sigma methodology to use, how to find hidden misunderstandings between people from different knowledge domains, and how to simulate the impact of improvements to iterative processes. All these methods and tools have been developed and tested at Siemens. This has been an action research project, where the author has been employed by the company under investigation for eleven years and has actively influenced the changes in the company based on knowledge gained at the company as well as on research studies conducted at universities. In action research the change initiative under investigation is conducted and analysed in a single context. The readers are invited to draw their own conclusions on the applicability of the results to their own specific cases. In addition to this, some conclusions derived using analytical generalisation, applicable to a more general case, are presented in the thesis. / <p>Defended att Chalmers University of Technolgy in 2007.</p>