Reservoir Computing (RC) is a leading-edge paradigm for the design and training of recurrent neural network models. The approach has become popular among practitioners due to its simplicity of implementation, effectiveness in applications, and efficiency. Theoretically, RC allows to deepen the study of characterizations and initialization of dynamical neural models, improving our understanding of their dynamical behavior. In recent years, studies on neuromorphic implementations of RC have also opened the way to breakthrough advancements enabling ultra-fast learning in the temporal domain. This special issue contains selected papers that provide an overview of the main streams of current research in the RC field.