Knowledge Space Theory (KST) and its Competence-based extensions (CbKST) are prominent examples on how an adaptation strategy for guidance in assessment and learning are grounded on a theoretical framework and can be applied in e-assessment and e-learning systems. KST and CbKST constitute sound formal frameworks for both structuring knowledge domains as well as for representing the knowledge and the competences of learners in detail. They are the basis for personalised assessment and individualised teaching. The development of this approach will be mentioned and illustrated by several applications, e.g. in terms of adaptive assessment and scaffolding. In it’s second part the talk will focus on two recent integrative developments, which have been performed within European R&D-projects and might be challenging for Large Scale Assessments (LSA): (a) Integrating Self-Regulated Learning (SRL) and Adaptive e-Learning Systems on the one hand and (b) creating Microadaptivity for e.g. Game-Based Learning (GBL) on the other hand. CbKST and SRL are rather contradictory, nevertheless they can be combined to a uniform and new approach taking advantage of each side. Microadaptivity in its current form combines Competence Structures (from CbKST) with Problem Spaces (from Information-Processing Theory of Human Problem Solving) for the purpose of non-invasive, process-oriented skill assessment, skill activation, skill acquisition, and motivation. The third part of the talk will summarise and reflect on challenges for LSA and Computerised Adaptive Testing (CAT) in face of current and future developments like Open Learning Environments, MOOCs, Cultural Diversity in Learning and Teaching, Inquiry-based Learning etc.