Is it true that, as the mainstream intuition asserts, urbanization and industrialization are the two main socioeconomic drivers of PM2.5? How do the two trends affect PM2.5 emission? This paper quantitatively analyzes the socioeconomic drivers of PM2.5 through assessment on Stochastic Impacts by Regression on Population, Affluence and Technology (STRIPAT), based on the panel data of 79 developing countries over 2001-2010. The average levels of PM2.5 pollution are calculated using remote sensing data, overcoming the difficulties that developing countries are in lack of PM2.5 monitors and that point data cannot reflect the overall level of PM2.5 pollution on a large scale. Squared terms of income and urbanization and their cross term are included in the regression models respectively to analyze the possible heterogeneous impacts on PM2.5 emissions in different development stages. The results show that income, urbanization and service sector have significant impact on PM2.5 pollution. Income has a positive effect on PM2.5 all the time but the effect decreases as the level of urbanization or income goes up. An inverted U relationship exists between urbanization and PM2.5, in which PM2.5 pollution positively correlates with a low level of income or urbanization but negatively at a high level. Policy recommendations from the perspective of macro-level social and economic regulation are provided for developing economies to reduce PM2.5 pollution.