This paper presents a fit between model and experiment for well-humidified polymer electrolyte fuel cells operated to maximum current density with a range of cathode gas compositions. The model considers, in detail, losses caused by : (i) interfacial kinetics at the Pt/ionomer interface, (ii) gas-transport and ionic-conductivity Limitations in the catalyst layer, and (iii) gas-transport limitations in the cathode backing. Our experimental data were collected with cells that utilized thin-film catalyst layers bonded directly to the membrane, and a separate catalyst-free hydrophobic backing layer. This structure allows a clearer resolution of the processes taking place in each of these distinguishable parts of the cathode. In our final comparison of model predictions with the experimental data, we stress the simultaneous fit of a family of complete polarization curves obtained for gas compositions ranging from 5 atm O-2 to a mixture of 5% O-2 in N-2, employing in each case the same model parameters for interfacial kinetics, catalyst-layer transport, and backing-layer transport. This approach allowed us to evaluate losses in the cathode backing and in the cathode catalyst layer, and thus identify the improvements required to enhance the performance of air cathodes in polymer electrolyte fuel cells. Finally, we show that effects of graded depletion in oxygen along the gas flow channel can be accurately modeled using a uniform effective oxygen concentration in the flow channel, equal to the average of inlet and exit concentrations. This approach has enabled simplified and accurate consideration of oxygen utilization effects.