Exams are central, quality-assuring elements in higher education that must meet high standards. In practice, however, this ideal is often not met. In this article, we outline an evidence-based procedure for the design and scoring of written exams and examinations that facilitates the handling of typical problems and reduces potential risks as much as possible, according to the current state of research. We discuss which quality criteria should be applied to university exams; which item formats can be used, with their specific advantages and disadvantages; how the guessing problem should be addressed in selected response test items; and the potentials of computer-based exams. Furthermore, we discuss the importance of different exam types to control students' learning behavior, the role of students' testwiseness, and the necessity of implementing systematic quality management for written exams. In addition, we aim to further deliberate on a constructive dialogue about challenges and the need for optimization as well as the prerequisites and conditions to construct better exams in university studies of psychology.