Among different technologies, dense ceramic membranes are prospective candidates for separation and purification of H2 in non-galvanic mode. By using integrated systems such as membrane reactors, in which both the reaction and separation are carried out in the same device, the system efficiency increases and makes scale-down economically efficient. Materials investigated for high-temperature hydrogen separation are divided into two classes: (i) single-phase materials and (ii) dual-phase materials, including cer–met and cer–cer composites. Moreover, mixed ionic–electronic conducting membranes are divided by their architecture into symmetric (self-supported) and asymmetric systems consisting of a dense ceramic supported on a porous substrate. Better understanding of the complex phenomena involved during H2 separation is necessary in order to tailor the materials composition. In fact, ceramic architectures based on dense mixed ion–electron conductors (MIECs) are extremely selective at medium–high temperatures (up to 900 °C) but do not reach yet the minimum-targeted flux of 30 mL min−1 cm−2 required to apply them in industrial processes. Nowadays, the best fluxes are of a few mL min−1 cm−2 and are obtained on dual-phase perovskites. Tubular membranes are technologically preferred respect to planar ones. With this aim, this chapter provides an overview of recent research on the most promising ceramics and ceramic composites and a reminder of the hydrogen-selective permeation mechanism in ceramic-based membranes.