Recent economic development in many Asian and European countries has shown an increase in vehicle kilometre travel (VKT) in many cities, which has resulted in an increase in the vehicular pollution levels. In particular, particulate matter (PM) concentrations emitted from vehicles are at alarming levels in most of the cities of the world. Therefore, there is growing interest in the formulation of local level air quality management system to tackle vehicular emissions. Many mathematical models have been widely used as tools in urban air quality management. In the present paper, an attempt has been made to evaluate the performance of the ADMS Urban model in predicting roadside PM concentrations at two cities namely Chennai, India and Newcastle, UK during critical winter period of the year 2009. The statistical parameters such as Index of Agreement (IA), Fractional Bias (FB), Normalized Mean Square Error (NMSE), Geometric Mean Bias (MG) and Geometric Mean Variance (VG) have been used to evaluate the ADMS model performance. Results indicated that the roadside PM concentrations predicted by the ADMS model are reasonably accurate for Newcastle than at the roadside in Chennai. © 2011 WIT Press.