Ceria has become an increasingly important component in automotive exhaust catalysts over the past decade. Recently, with the proposal that measurements of oxygen storage be used for the on-board evaluation of catalyst performance for both low emission vehicles (LEV) and non-LEV vehicles, understanding the role of ceria and its deterioration with catalyst aging has become even more important. It is well established that ceria in an alumina support promotes oxygen storage/release by automotive catalysts under cycled air/fuel conditions, which in turn promotes the catalyst's conversion performance under those conditions. Another benefit of ceria is its enhancement of the catalytic activity for other reactions, such as the water-gas shift reaction under rich conditions. In addition, ceria may help catalyst durability by promoting precious metal dispersion and playing some role as a stabilizer of the support.To examine, in part, how these benefits of ceria persist in aged catalysts, we have carried out emission measurements on catalysts with varying cerium levels (0-50%) that have received increasing amounts of accelerated aging. We find that for the washcoat formulation investigated, catalysts with a ceria content of about 30% provide the highest conversion efficiencies for HC, CO, and NOx after extensive high temperature treatments designed to mimic up to 100,000 miles of vehicle aging. We have also measured conversion efficiencies for catalysts, including warm-up catalysts in front of the underfloor location, which have been exposed to high temperatures and suffered loss of oxygen storage capacity, as measured using a multiple oxygen sensor approach to meet OBD-II regulations. These catalysts still provide high conversion efficiencies on powertrains which have well controlled A/F ratios. Presumably the oxygen storage capacity of ceria can decrease after high temperature exposure, due to the decreased surface area of the ceria and the precious metal particles and their decreased interaction with each other, while the precious metal surface area is still sufficient for good performance. To the extent that there is little correlation between the oxygen storage/release capacity of a catalyst system and its HC conversion activity, methods that use oxygen storage to determine catalyst activity, particularly for warm-up catalysts, will not provide the level of accuracy required for reliable on-board monitoring of catalyst performance.