The literature reported different studies of soot combustion reaction under very distinct experimental conditions, which can include different values of catalyst:soot weight ratios, gas flow and heating rates. Therefore, avoiding screening of innumerable catalysts or empirical experiments, this work aims to present a general methodology based on a statistical experimental design of experiments with soot combustion, evaluating different reaction conditions and parameters that can be used for any other similar study. In this way, the effect of experimental conditions on the parameters used for evaluating the performance of Mo/Al2O3, a promising system previously studied, and Pt/Al2O3, a notorious catalytic system, were studied by a complete factorial experimental design. The results have shown that the experimental conditions strongly interfere with the parameters used for evaluating the catalytic performance and then it may generate incorrect conclusions. The effects of interaction between different conditions on the activity and mainly on the selectivity of CO2 permitted to explain the performance of catalysts on soot combustion and to distinguish different pathways of catalytic and non-catalytic reactions under specific reaction conditions. The most appropriate conditions for studying soot combustion seem to be high cat:soot ratios, low heating rates and high gas flow rates, which, according to this work, must be equal to: 95:1,2 K min(-1) and 115 mL min(-1), respectively. (C) 2008 Elsevier B.V. All rights reserved.