The Milestones Project, like all previous systems and changes in graduate psychiatric education, for example, moving from 3 to 4 years of training or adopting six competency domains, has been devised without any supporting data and does not assess meaningful outcomes, such as improved patient outcomes. No evidence is presented that Milestones-based training will produce better psychiatrists. There is a path forward. First, replace unproven expert consensus with scientific and evidence-based approaches. Second, exchange endpoints that are easy to assess but uncorrelated with real world functioning (e.g., multiple-choice examinations) for outcomes that are meaningful and external to the training program (e.g., patient outcomes). Finally, to prevent possible waste, excess burden, or harm, no changes should be mandated until proven in prospective studies.