Approaches to models and computer simulations of conductivity, polarization resistance, and impedance of composite electrodes in solid oxide fuel cells (SOFC) are reviewed with respect to the more important experimental findings. The approaches are classified according to how they model the highly disordered structure of composite SOFC electrodes: As corrugated layers of electrode material covered by a thin film of electrolyte or vice versa (thin film model), as a random packing of particles (Monte Carlo calculations), or using a macroscopic, averaged description of the disordered electrode structure (macroscopic porous-electrode model). Thin film models appear to be useful rationalizations of some experimental measurements of polarization resistance, but in the stricter sense fail to predict a number of important electrode characteristics. The Monte Carlo method, on the other hand, apparently meets with most of the more prominent experimental results reported so far, although some issues concerning parameter choices, among other things, remain to be resolved. The macroscopic porous-electrode theory may serve as a useful simplification of the Monte Carlo method, but with a more limited scope. Modeling of composite electrodes for SOFC thus appears to have reached a level where it can be used for practical engineering applications. As an example of this, the rate of methane reforming at Ni-YSZ cermet anodes under current load is calculated using the framework of the macroscopic porous-electrode theory, modified to include non-linear kinetics and gas-phase diffusion. The reforming reaction is quite evenly distributed in the anode, and its overall rate is therefore strongly dependent on thickness. However, most of the electrochemical reaction is likely to occur in a region closer than 10 μm to the bulk electrolyte. For an anode thickness larger than this, the current-collector potential at a given current is by and large independent of thickness. The ratio between the rates of the reforming and the electrochemical reactions can therefore be balanced to a certain degree by optimizing thickness, without significant loss in cell power. In addition, cermet porosity, volume fraction of Ni and Ni-particle size, appears to have a moderate effect in controlling the rate balance, which will have to be manipulated within the constraints set by the requirement of percolation in the gas-phase and the Ni- and YSZ-networks.