Finding the best match between learners' expertise and instruction is a central issue in educational psychology. The idea that different learners might need different instruction gave rise to the concept of aptitude-treatment interaction (ATI) launched by Cronbach and Snow (1977). The expertise reversal effect is a variant of an ATI effect. It occurs, when an instructional format that is beneficial for novices compared to other formats looses its advantage with increasing expertise of the learners and finally becomes disadvantageous for individuals with higher expertise (Kalyuga et al. 2003).Expertise reversal effects have been found for example by Kalyuga et al. (1998Kalyuga et al. ( , 2000Kalyuga et al. ( , 2001aKalyuga et al. ( , b, 2003, Leahy et al. (2003). The contributions of this special issue further elaborate these findings. Oksa, Kalyuga, and Chandler (this issue) demonstrate that the expertise reversal effect can be found not only in well-structured, but also in ill-structured domains such as the interpretation of literacy. Nückles, Hübner, Dümer, and Renkl (this issue) show that the expertise reversal effect has also motivational aspects and applies to the use of strategies. The article of Homer and Plass (this issue) indicates that the effect is not only related to domain-specific prior knowledge, but also to the general developmental level of learners. The findings of Blayney, Kalyuga, and Sweller (this issue) argue for the importance of adapting learning environments to the changing levels of learner expertise in the use of spreadsheets. Finally, Salden, Aleven, Schwonke, and Renkl (this issue) address the question of balance between providing sufficient instructional assistance on the one hand and self-regulated generative learning on the other hand in the case of worked examples, and they demonstrate that adaptive fading of instructional assistance is crucial for effective learning.There are some fundamental assumptions that can be found throughout the contributions of this special issue and in other papers on cognitive load theory. One assumption is that cognitive load is contingent to the limited capacity and duration of working memory. Another assumption refers to the concept of redundancy. It implies that information that is necessary for novices may become redundant for more advanced learners and may overload working memory. The following comments will analyze these assumptions more