Abstract: In air quality management systems, aiming at the existence of multiple uncertainties and system complexities, based on the methods of interval linear programming (ILP), stochastic mathematical programming (SMP) and fuzzy possibilistic programming (FPP), an interval stochastic fuzzy possibilistic programming (ISFPP) model was developed to identify effective management policies. The developed ISFPP model can not only deal with multiple uncertainties, but also reflect system complexities. Moreover, the ISFPP model can help to analyze various management scenarios associated with different confidence levels. The ISFPP model was applied to a hypothetical case study of air quality management. The results indicate that the change of confidence level may lead to the corresponding changes of total system cost, treatment amounts of pollutants and emission amounts exceeding the standards; moreover, under different confidence levels, different enterprises can select the appropriate pollutant control measures and determine the reasonable treatment amounts of pollutants and emission amounts exceeding the standards. Thus, the modeling results can be used for generating decision alternatives, and help the decision-makers to identify desired management policies.