Which form of exchange governance performs better in a dynamic environment? This remains an unanswered question in the transaction cost analysis (TCA) and relational exchange literatures. Some researchers purport that transactional governance provides superior performance by providing firms the flexibility to change suppliers. Others suggest that relational governance leads to superior performance because of the willingness of both parties to adapt. Reviews of TCA have turned up ambivalent empirical findings with regard to the effects of uncertainty despite a track record of strong empirical support for other predictions. Because most of TCA and relational exchange theories' predictions enjoy strong support, this research builds upon these theories to propose a theoretical modeling framework for a dynamic environment in a supply chain network (SCN) setting. This dissertation extends TCA and relational exchange to a dynamic, network environment. It uses the approach of building a simulation in order to study in detail the relationship between key exchange factors and the selection of transactional and relational exchange governance over time. This research effort extended TCA theory with a complex adaptive model of supply chain network governance evolution that attempts to link environmental, network, production, firm and exchange factors in a continuously evolving loop. The proposed framework expands transaction cost analysis' explanatory power. Results partially support past scholarly proposal that uncertainty functions as an antecedent of asset specificity rather than as an independent construct affecting governance outcome dependent upon which form of uncertainty is being considered. The successful simulation of supply chain networks as complex adaptive systems shift the focus from deterministic, confirmatory models of exchange to an exploratory, positive model. Instead of exchange governance as an outcome, it is the catalyst of the evolutionary process.