This commentary describes a program developed to evaluate patients with undiagnosed and rare diseases as a way to reduce diagnostic error. The authors explain heuristic and clinical reasoning shortcomings that can hinder diagnosis and how to manage them through improved information gathering and hypothesis testing.