From the 1930s onwards, the problem of the transition from primary to secondary school has been a source of discussion and contradiction in the Netherlands. For many years it has been disputed whether this transition problem should be conceived as a problem of selection (testing) or as an educational problem (reforming the instruction and curriculum). The psychologist Kohnstamm (et al.,) did, with respect to the transition question, relevant research on students' learning, thinking and learning to think. He considered the psychological investigations on the development of human thinking by the German Schools as fruitful inspiration, for instance: the Würzburger School (Külpe) and the Mannheimer School (Selz). Both schools criticized the influential, sensualistic theory of association advocated by Herbart, a theory that was popular around 1900. Both schools believed that (insightful) thinking of students is based on the activity of the students and not merely on associations as in a mechanical process. Introspection was a suitable research method, they asserted, to investigate these thinking activities. Külpe argues that thinking is connected with the activity of the individual, triggered by the demands of the task, and hence not just by laws of association. Selz argues that thinking subjects strive for filling up a structure (complex), and this process can be interpreted as the construction of a meaningful Gestalt. Subjects do not react to a problem in an associative way, but in a process Selz characterizes as an 'anticipating scheme'. Higher order thinking abilities can be developed and improved, according to Selz, in education. After more than a century the discussion reiterates. Some authors tell us that drill and practice offers a primary and solid basis for insightful learning, while others claim that meaningful, interactive and reflective learning are the core processes. Students should, after all, understand how and why they act the way they do. Kohnstamm conceived the transition problem, at first, as a selection (test) problem, but in later years he saw it more and more as an educational problem. Moreover, he recommended, it is useful to construct scholastic tasks, while school directors should be trained in composing reliable reports about the capacities of the students. He preferred elaborating qualitative assessments of thinking processes instead of using merely static, quantitative rankings (Gauss). After a century of discussion and educational research, the problem of the transition has not, in many important aspects, been solved. We are still confronted, and probably not only in the Netherlands, with serious selection and transition problems with all the consequences for educational equity and equal chances in education for especially minority groups in our society.