In this paper, educational and technical challenges for applying learning pathways in Massive(ly) Open Online Courses in higher education are outlined. We argue that quality issues and didactical concerns may be overcome by(1) reverting to small Open Educational Resources that are (2) adaptively joined into concise courses by considering (3) predefined learning pathways with proper semantic annotations and (4) the observation of learner behaviour. Such a merger does not only require conceptual work and corresponding support tools, but also a new meta data format and an engine which interprets the semantic annotations as well as the measures of learner’s actions. These factors are then turned into didactically meaningful recommendations for the next learning steps, thereby creating a personalized learning pathway for each learner. The EU FP7 project INTUITEL is introduced, which has already contributed to the conceptual work and is currently developing the software to achieve these tasks.