The Westland-Oostland Greenport is a large Dutch industrial cluster that is economically, socially and culturally important. Like many industries and regions, the greenport is facing serious challenges as a result of globalisation, energy supplies, and greenhouse gas emissions. These challenges can be understood as sustainability issues that come from an imbalance between the economic and social values that the greenport presently enjoys and the environmental values that it lacks. Greenports are complex adaptive systems, which means that the different challenges and values are all interconnected in complex and dynamic ways. This also means that improving the greenport is a matter of helping it evolve from its current state of imbalance toward a new state of sustainable balance that will help the greenport endure long into the future. The problems of unsustainability are growing increasingly urgent, which means the greenport can hardly afford to wait for a spontaneous change toward sustainability. Unfortunately, complex adaptive systems and state changes within those systems cannot be managed top-down. Thus, the Westland-Oostland Greenport wants to know `How can we manage a transition to sustainability?' In response, the greenport is interested in Transition Management (TM), an approach to scientific research and policy formation that proposes to steer complex adaptive systems toward sustainability through a blend of bottom-up and top-down measures. TM certainly looks promising, but it is not yet clear ``what is the scope or feasibility of managing transitions''. On behalf of the Westland-Oostland Greenport, this work uses TM to explore how a transition to sustainability might be managed within in the Westland-Oostland Greenport. At the same time, this work also reflects on TM itself by questioning whether sustainable transitions can be managed at all. The Research Questions There are two main research questions, each with two sub-questions. The first main research question is ``Can TM help the Westland-Oostland Greenport manage a transition to sustainability?'' which can be broken down into the following sub-questions: What new understanding and insight can be gained from applying TM to the Westland-Oostland Greenport? What influence, policy recommendations and practical advice can be derived from the new understanding and insights gained by applying TM approach to the Westland-Oostland Greenport? The second main research question is ``How can TM be improved as a consequence of being applied in the Westland-Oostland Greenport?'' which has the following sub-questions: What potentially problematic assumptions can be seen within the TM approach as it was applied to the Westland-Oostland Greenport? What insights, further questions, or improvements for TM come from exploring these problematic assumptions? The Research Methods The research begins with a literature review to explore the most important ideas of complex adaptive systems, sustainability and TM. The first main research question is then addressed by applying several research methods to the Westland-Oostland Greenport, all of which are consistent with TM. These research methods include a case study, a participatory workshop, and an agent-based model. Next, the second main research question is addressed through a series of relatively abstract agent-based models and modelling experiments. Each model is designed to test an assumption within TM about how complex adaptive systems work or about how transitions can be managed. Results Part I covers the application of various research methods to the Westland-Oostland Greenport. Each of these produces a set of new insights into the greenport and its sustainability problems. On the whole, the insights tend to relate to specifics, such as the identification of specific drivers behind a diffusion of interest, specific gaps in local sustainability policies, and specific stakeholder features that influence relevant behaviour. Many of these insights are turned into practical recommendations that policy-makers could use to help manage a transition to sustainability within the greenport. For example, recognising the specific drivers behind a particularly rapid and important technology diffusion can help policy-makers drive desirable diffusions in the future. Similarly, identifying gaps in current sustainability policies allows policy-makers to develop more effective programmes without duplicating existing programmes. Not all of the insights are easily converted to direct policy recommendations; some are better understood as advice on the process and typical problems associated with policy formation. For example, recognising that innovators are quick to embrace change is usually seen as a good thing when policy-makers want to encourage a desirable change, but quickly becomes problematic after a desired change has been effected and further change is seen as less desirable changes. Some specific insights could not have been predicted, but none were inconsistent with general TM expectations. For example, energy market liberalisation proved to be one of the most important drivers of the combined heat and power technology diffusion, even though the technology diffusion was not an intended or expected consequence of long term national energy market policies. TM expects transitions and diffusions to occur when `windows of opportunity' are opened. Thus, even though this particular diffusion was not an intended result of its most important driver, neither was it an especially surprising outcome. In this way, the policy recommendations based on the insights uncovered in Part I are all consistent with the theory of TM and with past TM policy recommendations. Part II covers the agent-based models that investigate various potentially problematic assumptions within TM. Each of these models produces new insights, some of which can be related to the Westland-Oostland Greenport but most of which relate to TM itself. None of the assumptions held up particularly well to investigation. As a result, and unlike the relatively cut-and-dried insights in Part I, the insights in Part II are best understood as calls for more critical analysis of TM, creative reinterpretation of past observations, and innovative approaches to understanding complex adaptive systems and sustainability. For example, the first of four agent-based models tested the commonly relied upon TM assumption that a large and diverse committee can produce a more objective system description that can a single individual. The results of that model suggested that all system definitions are mutually exclusive of all others, regardless of how they were produced, suggesting that no system description is obviously more objective than any other. Another agent-based model examined the pervasive TM assumption that innovation and selection are opposing forces but found no evidence to support the proposed innovation-selection relationship nor the many TM programmes that rely on the assumed relationship. Importantly, these insight do not claim to offer a definitive answer or superior explanation to replace the TM assumptions examined by the models. For example, the conclusion that diverse committees do not produce more objective system definitions but does not entail any proposal for an alternative process or structure that produces more objective system definitions. Likewise, refuting the assumed relationship between innovation and selection is not the same as proving that innovation and selection have some other relationship instead, although an alternative view of innovation and selection is offered as a possibility for further investigation. As a consequence of the different research questions and approaches used, the policy recommendations and insights from Part II are less practical and specific than those from Part I. For example, policy-makers are not advised to avoid large and diverse committees when creating system descriptions, but are urged to focus on the actual benefits of wider public engagement rather than unsubstantiated assumptions about objectivity. Also unlike Part I, some of the insight and recommendations of Part II are aimed at users or proponents of TM rather than directly at policy-makers. For example, many TM practices rely on using (more) objective system descriptions. By undermining the assumed objectivity of system descriptions, the experimental results call the validity of all of these practices into question. When the results of all four of the various agent based modelling experiments are taken together, serious inconsistencies become visible within TM, some of its assumptions begin to appear very flawed, and the whole field looks as if it would benefit from some deep and critical self-reflection. Conclusions Overall, the works finds that applying TM research methods to a complex adaptive system generates new insights about the specific details of that systems, much of which can be used to create practical recommendations or justify policies in relation to sustainability efforts. The new insights and details revealed by the research may be surprising in their uniqueness or specificity, showing that TM research can be extremely valuable. Despite their potentially surprising uniqueness or specificity, TM research methods are unlikely to produce any results, insights or policy guidance that are truly unexpected or that are inconsistent with established TM theory or existing policy proposals. At the same time, it is important to note that TM has not yet managed a transition to sustainability. Perhaps TM merely needs more time to achieve the desired outcome, but it is also possible that more and more of the same TM efforts will only produce more and more of the same lack of success. The work also finds that TM, as it currently stands, contains some incoherent ideas and relies on some unsupported assumptions. These inconsistencies and assumptions may be hindering TM efforts to achieve sustainability, so TM is encouraged to critically reflect on its ideas about complex adaptive systems, transitions and sustainability as well as its own processes, research methods and sustainability efforts. In so doing, TM may correct some problems, discover new and better ways to work, and become a more effective tool for managing systems and moving them toward sustainability. This critical self-analysis may also reveal that TM is fundamentally flawed and should be discarded, but that too could inspiring researchers to devise entirely new approaches that are more successful in achieving sustainability.