The aim of this paper is to argue for a number of statements about what is important for a client to do in order to improve quality in new infrastructure projects, with a focus on procurement and organizational issues. The paper synthesizes theoretical and empirical results concerning organizational performance, especially the role of the client for the quality of a project. The theoretical framework used is contract theory and transaction cost theory, where assumptions about rationality and self-interest are made and where incentive problems, asymmetric information, and moral hazard are central concepts. It is argued that choice of procurement type will not be a crucial factor. There is no procurement method that guarantees a better quality than another. We argue that given the right conditions all procurement methods can give good results, and given the wrong conditions, all of them can lead to low quality. What is crucial is how the client organization manages knowledge and the incentives for the members of the organization. This can be summarized as "organizational culture." One way to improve knowledge and create incentives is to use independent second opinions in a systematic way.