A series of specific choices, made by both the antagonists in a conflict and the third parties that mediate, determine the distribution of mediation resources. These choices, which include the selection of whether to have mediation, the choice of who mediates, and the adoption of particular mediator strategies, ultimately affect the prospects for successful outcomes. Studying the mediation selection process has two primary purposes. First, it prevents misleading inferences caused by the involved actors conditioning their mediation preferences on their expectations of the outcomes. Second, it clarifies the constraints on optimal mediation, and thus provides insight into how effective mediation can be more prevalent. Using quantitative analyses, the results demonstrate that mediation occurs when the bargaining problems are manageable enough for success to be likely, or when there are conditions that give the actors devious incentives for mediation. In the cases that receive mediation, the conflict actors prefer the most potentially effective intermediaries when the bargaining problems are less manageable. The project also finds that when persuasive third parties derive sufficient benefits, mediation is more likely to occur, more likely to involve these types of third parties, and more likely to include the most costly tactics. Taking steps to mitigate selection and omitted variable biases substantially affects the inferences drawn regarding what contributes to effective mediation. Factors such as the type of mediator, power balance, changes in power balance, previous interaction, ethnic dimensions and overall third-party incentives to mediate appear to have no effect on mediation outcomes in analyses that do not take into account the selection process. Once the models are informed by the information in the selection process, the important effects of these factors materialize. Accounting for the selection process, the findings demonstrate that mediation is an effective vehicle toward peaceful resolution. In addition, the findings reveal that mediation performs best when the most resourceful mediators are willing to use all the tactics at their disposal, and when the belligerents do not have incentives to use mediation merely as a stalling tactic