The main objectives of the research were to estimate the relationship between the gas-particle partition coefficient, K p , and the octanol-air partition coefficient, K OA , of polycyclic aromatic hydrocarbons, PAHs, and polychlorinated biphenyls, PCBs, at industrial and urban sites in the Vojvodina region, to compare the obtained slopes and intercepts of the log K p vs. log K OA relations with the results of regression analyses reported in previous studies and to assess the consistency between the particle-bound fractions predicted by the K OA absorption model and the results obtained within field measurements. Fairly good log K P -log K OA correlations, with an average value of the correlation coefficients of 0.70, indicate that the partition coefficient K OA can be used as a prediction parameter of the gas-particle partitioning processes for both classes of compounds. The results of modelling the atmospheric distribution of PAHs using the K OA absorption model showed inconsistencies between the measured and predicted values of the particle-bound fraction, φ, of 1-2 orders of magnitude, while significantly higher discrepancies for PAHs in the Nap-Ace range were found. A similar variability of the measured/modelled φ values was obtained using the Junge-Pankow adsorption model, indicating the presence of particles enriched with PAHs. The conducted research showed that the K OA -based approach was less suitable for predicting the gas-particle partitioning of PCBs in urban and industrial sites, compared to the Junge-Pankow model.