In the present paper, the effect of sulfur on the performance of the catalyst particle and the operation of an industrial side-fired reformer is analyzed, by coupling detailed mathematical models for the reactor and the sulfur poisoning. The model predicts the transient evolution of the main variables inside the catalyst pellet (e.g., poison coverages and reaction rates) and along the axial position (e.g., gas composition, sulfur distribution, gas and tube skin temperatures). The common simplified model, which neglects the reaction rates update as the poisoning proceeds, is compared with the model proposed in this article. The deactivation phenomenon causes methane reaction rates that present maxima along the catalyst coordinate, moreover effectiveness factors higher than the unity are found close to the reactor entrance. When the poison is removed from the feedstock, non-monotonic sulfur distributions appear inside the particle and simultaneous sulfur migration towards the surface and the center of the pellet is observed. Non-isothermal deactivation conditions similar to those found in industrial reformers are simulated, i.e., the heat input is assumed constant during the poisoning. The increments in the gas and tube skin temperatures are maximum at those axial positions where the observed methane reaction rate becomes equal to its value at conditions of fresh catalyst. The proposed mathematical model is a useful tool to describe the dynamic behavior of the macroscopic reactor variables in terms of the local phenomena that take place within the catalyst particles. In practice, this model can be utilized to evaluate key variables such as: the outlet sulfur content (to minimize poisoning downstream the reformer), the maximum tube skin temperature (to extend the tube-life time) and the axial distribution of the process gas temperature and composition (to prevent carbon formation in the top section of the reactor).