This article aims at contributing to the extant literature on government make-or-buy choices building on transaction cost economics (TCE) by explicitly theorizing about transaction alignment and its relation to performance. It is argued that current theoretical and empirical models of government make-or-buy choices are not able to make predictions that corroborate theory. They are depended upon the assumption of perfect competition that is ill-suited for the public sector setting. Instead contingent models that take performance differences into account constitute a more valid model of TCE in this setting. In this article theoretical models and empirical approaches for such a research agenda are developed.