Abstract:This article reports findings from an empirical study of Settlement Week mediation in four general civil trial courts. Nearly all parties and lawyers had favorable assessments of the mediation process, the mediator and, in cases that settled, the agreement. The only mediator qualification that increased the likelihood of settlement was the mediators' level of mediation experience. The mediators' hours of training, years in law practice, and expertise in the subject matter of the case were not related to settlement. Although the mediators' recommending a particular resolution increased the likelihood of settlement, it did so at the expense of parties' perceptions of fairness. Findings regarding the effect of other mediator qualifications and actions on settlement and on participants' assessments are also discussed.