Changes in the combination of air which can cause deterioration for human health or environmental balances, or the mixing of substances in air that should not be in it causes air pollution. The air layer has been polluted by the wastes generated during production and consumption activities resulting from various activities of humans and this pollution adversely effects living life on the earth. The aim of this study is to examine the factors affecting air pollution by applying a cluster analysis to mixed variable panel data. In line with this purpose, a mixed variable panel dataset has been designed for 28 countries from Eurostat for the period 2007-2013. In this data set, air pollution have been determined as dependent variable. Industrial production, greenhouse gas emissions, garbage waste, investment in the environment and whether the countries are members of the European Union or not have been determined as independent variables of this data set. The cluster analysis has been applied to the panel data set by using Ward's method with Gower distance and proposed distance. For each cluster obtained, cross-sectional dependence, heteroscedasticity and autocorrelation assumptions have been tested and deviations from the assumptions have been observed. Therefore, Parks-Kmenta (GLS), Beck-Katz (PCSE) and Driscoll-Kraay estimators which are robust to the deviations from the assumptions, have been applied for parameter estimation. As a result of the study, it has been seen that the factors affecting air pollution vary according to clusters.