Abstract Air pollution is a common problem in mega-cities in Asia, resulting in high levels of particulate matter (PM). In this study, 24 h samples of airborne particulate matter (PM 2.5 and PM 10 ) was collected for two weeks in urban locations in Kabul and Mazar-e Sharif, Afghanistan. The samples were analyzed for twenty n -alkanes, fifteen polycyclic aromatic hydrocarbons (PAHs) and eighteen oxygenated PAHs (oxy-PAHs). The results were compared to a reference location in Umea, Sweden. The main objectives were: to characterize the organic fraction in PM 2.5 , with focus on oxy-PAHs, PAHs and n -alkanes, assess relationships between compound groups, and investigate possible source categories. Mean PM 10 , levels were 260 μg m −3 , and 334 μg m −3 and mean PM 2.5 levels were 86 μg/m 3 and 68 μg m −3 in Kabul and Mazar-e Sharif, respectively. The higher ratio of PM 2.5 to PM 10 in Kabul than Mazar-e Sharif or Umea was reflected in the high number of small particles, as collected by a particle counter. Kabul and Mazar-e Sharif had high levels of PAHs and oxy-PAHs compared to Umea and previously reported data from Western countries. Generally, the highest individual levels of oxy-PAHs were found for ketones (0.27–33 ng m −3 ), whereas the quinone levels were lower (0.027–3.1 ng m −3 ). High correlations ( r > 0.96) between PAHs and oxy-PAHs were found at both locations. Diagnostics for n -alkanes, such as the carbon preference index, and the most abundant species, C max , and the presence of unresolved complex mixture indicated a higher contribution from coal and petroleum sources in Kabul. Models generated by principal component analysis (PCA) and positive matrix factorization (PMF) both suggested separate sources for lower molecular weight n -alkanes and higher molecular weight n -alkanes. The PAHs and the oxy-PAHs were not separated in either the PCA or the PMF models, indicating that they had similar sources and are perhaps relatively unspecific as source markers. Nevertheless, the PAHs and oxy-PAHs were good markers for bad air quality in the urban locations in Afghanistan in comparison to data from Western countries. Based on the models and diagnostic ratios it is suggested that traffic and combustion of coal and biomass were the dominating sources of the high levels of particle-bound organic compounds.