Computer simulation is now considered to be a crucial stage in the design of automotive catalysts due to the increasing complexity of modern aftertreatment systems. The resulting models almost invariably include surface reaction kinetics that are measured under controlled conditions similar to those found on a vehicle. Repeatability of the measurements used to infer surface reaction rates is fundamental to the accuracy of the resulting catalyst model. To achieve the required level of repeatability, it is necessary to ensure that the catalyst sample in question is stable and that its activity does not change during the test phase. It is therefore essential that the catalyst has been lightly aged, or "de-greened" before testing begins. It is also known that the state of the catalyst's surface prior to testing has an impact on its subsequent light-off performance and that test history can play an important role in catalyst activity. Suitably pre-treating the catalyst surface can ensure that a reference point is reached prior to a light-off test. The work summarized in this paper includes a study of both the de-greening phase and the state of the catalyst's surface on its activity, with the aim of developing a robust test protocol that provides repeatable kinetic data under realistic operating conditions.To establish a protocol for initial stabilization of new catalyst samples, a series of tests were conducted on catalysts that had been thermally aged at 600°C and 750°C respectively until the sample exhibited stability. The activity of the sample was assessed by repeating identical CO light-off tests following each period of time in the oven. A period of 8 hours at 750°C was found to sufficiently stabilize the sample.In the pre-treatment study, an investigation into the effect of several pre-treatment protocols on CO light-off was carried out on a three-way catalyst of commercial formulation in order to establish a robust protocol to ensure test repeatability. The investigation focused on pre-treating the catalyst sample in hydrogen, oxygen or nitrogen environments at specified conditions of flow and temperature. Following pre-treatment, identical CO light-off tests were performed in order to assess the respective influences of each pre-treatment. A trend of lower light-off temperatures was observed for all pre-treatment strategies on the first of two consecutive light-off temperature ramps with a notable increase in light-off temperature observed on the second ramp. As a result of this study a pre-treatment protocol was established. The protocol involves treating the catalyst sample in flowing nitrogen up to 600°C using a temperature ramp of 15°C/min.