Infrared thermography measurements in various oxidation reactions revealed high-temperature domains whose boundaries are either stationary, oscillating, moving or rotating. These motions were observed on catalytic wires, rings, cylindrical pellets and thin catalytic beds. Their evolutions are different from the classical Turing mechanism, which explains many of the pattern formation in reaction–diffusion system. The observed patterns are strongly affected by the interaction between the local surface reaction rate, the mixing in the surrounding reactant phase and by the intrinsic heterogeneity of the catalytic surface and the transport coefficients. While simulations can qualitatively predict such patterns, quantitative predictions require a reliable kinetic model. Formation of hot regions in a three-dimensional fixed-bed can be predicted with reliable kinetic models. While formation of hot regions in the flow direction in the reactor are reasonably well understood, there is still a need to gain qualitative understanding as well as design criteria for formation of hot zones transversal to the flow direction.