Abstract Crystallographic orientation relationships (CORs) of next-neighbour crystals represent a special case of crystallographic preferred orientation (CPO), where relative crystallographic orientations of neighbour-crystals follow defined rules of misorientation systematics (COR rules). The presence/absence and nature of crystallographic orientation relationships between next-neighbour crystals can be used to infer petrogenetic information from polycrystalline materials provided that the processes of COR formation are understood and parameters that control the kinetics of COR formation can be identified. After giving an overview on COR terminology, this chapter highlights non-genetic criteria for COR characterization, including a discussion of analytical methods that are used to constrain these criteria. The development of electron backscatter diffraction (EBSd) in scanning electron microscopy (SEM) has provided new information on CORs, which is complementary to data obtained from transmission electron microscopy (TEM) analysis. Based on these non-genetic criteria, different types of CORs are characterized. Subsequently, physical parameters that can potentially influence COR formation are discussed. Furthermore, different scenarios and mechanisms leading to COR formation are outlined together with examples from experiments and from natural mineral and rock systems. The different boundary conditions of COR formation in various petrogenetic scenarios and the potential mechanisms that have to be taken into account when studying COR genesis are addressed. This chapter highlights the necessity of a multi-stage investigative approach in COR studies. First, the presence/absence and nature of CORs needs to be analysed based on non-genetic criteria. In a second step the formation mechanism of the CORs under consideration must be constrained, before in a third step, petrogenetic information can potentially be inferred. Moving from the second to the third step requires understanding of the parameters controlling COR development, which is by no means complete and leaves open tasks for future COR research.