This Article contributes to recent scholarship regarding Long Term Agreements (“LTAs”) by providing empirical evidence that suppliers are more likely to undertake the costs of an LTA if the transaction requires significant capital expenditures or the potential for large sunk costs. Through a survey of a random group of sixty-three Ohio manufacturers, the Article explores why manufacturers with a full range of contractual and non-contractual solutions might choose one set of arrangements over others.1 It then seeks to link its findings to a broader theory of how parties bargain to solve durable problems under conditions of uncertainty, sunk costs, and opportunism, while minimizing costs. Although only a small portion (seventeen percent) of our sample size indicated that they used LTAs in the majority of their transactions, this group indicated they were more likely to produce customizable goods and have significant capital expenditures. Such a finding is consistent with a model of bargaining in which parties in a transaction seek to achieve their overall goals of wealth maximization while minimizing costs under conditions that include bounded rationality, sunk costs, and opportunism. If a product is customized for a particular buyer, and the supplier invests sunk costs toward customization, that investment makes it difficult and costly to exit the relationship or resell to others. Where such vulnerabilities exist, the need for protection may justify the costs of LTAs. The non-adoption of LTAs by some suppliers demonstrates that the new organizational form of networked firms, governed by an LTA and straddling markets and hierarchies, has not captured all of manufacturing and reflects a diversity of arrangement. The nonadoption of LTAs may be one way suppliers respond to the stresses and frictions of the new architecture of supplier relations. Those stresses show that the new organizational paradigm is not static and suffers from the same hazards as an exchange relation. The willingness of suppliers to adopt an LTA when facing large sunk costs shows the continuing importance of sunk costs in institutional decision making and offers an additional reason beyond the need to collaborate under conditions of uncertainty to explain why parties adopt LTAs. The other type of risks — opportunism and vulnerability from investing large resources — may be best handled by entering into an LTA because it offers security, including implicit protections needed for the supplier to invest. The switching costs that lock parties into a mutual dependency and protect parties who have invested comes gradually, but without the LTA, the supplier would be reluctant to undertake the initial investment. The importance of sunk costs may also explain the choice of buyers to operate under an LTA. Since many of the benefits of LTAs, including information sharing, could be achieved by buyers hierarchically and imposed on suppliers, the explanation for adopting LTAs may lie with the need to collaborate under conditions of uncertainty and the benefits in terms of added value derived from “managerial contracting” practices, but with the need to protect large investments through the security offered by an LTA. Thus, there are two functions of LTAs: (1) how-to provisions to guide and improve production; and (2) provisions offering security of a continuing commitment either through express provisions or implicit protections. This Article suggests that although information-sharing protocols serve to “institutionalize learning,” help parties when there is an “inability” to know how to solve a production problem, and offer more information to informally enforce new types of behavior that are non-compliant, these benefits might occur by means other than an LTA. For example, a quality manual may impose a quality assessment be done by the buyer at the supplier’s plant. Alternate means of obtaining the information outside of an LTA raise the question of why LTAs are adopted.