When states face an international cooperation problem requiring enforcement, when do they decide to make that enforcement formal versus informal? I introduce a research design for investigating how informal mechanisms might be relevant to formal international agreements. I present an overall theory of punishment provisions and a set of hypotheses about whether any needed punishments will be formalized or not. This theory gives rise to a two-part empirical analysis conducted on a large-n dataset. First, the presence of enforcement mechanisms in agreements is predicted, and, second, those cases that are “misclassified”aEuro”ones in which the model predicts the presence of such mechanisms, but the agreements lack them-are analyzed. These misclassified agreements, candidates for informal enforcement, are characterized by regime heterogeneity and military asymmetries among parties. Case study evidence supports the results.