One of the justifications for research on translation processes is that it will be of use in the training of translators. Very little work, however, has been done on the actual ways in which research can be used in training. This article explores the most direct mode of application, where actual process experiments are carried out in the classroom. Three such experiments are reported on: one on the use of online machine translation, another on translator styles, and a third on the effects of time pressure. The quantitative and qualitative results of the experiments indicate that this approach can be highly effective in stimulating students to find out about their own translating, and rather less than effective as a means of producing valid research findings. The direct use of research in the classroom should thus be considered of qualitative interest to the individual student rather than quantitatively valid as a way of producing knowledge of the general.