The Serbian E-PRTR register (The European Pollutant Release and Transfer Register) was established in 2007 and harmonized with PRTR protocol of Arhus convention and E-PRTR directive. In this paper, the quality of 2009 data reported to the Serbian PRTR register was analyzed. The analysis includes 37 large combustion plants with the capacity equal or greater than 50 MW. The combustion plants include power plants, heating plants and industrial energy units. The calculation is done using EMEP/EEA 2009 methodology and Tier 2 approach. The analysis obtained presents an overview of the quality of the reported data for SOx, NOx and TSP emissions for 61 combustion units e.g. emitters (stacks). The results show that all 61 emitters reported data with the deviation greater than 25%, with 55% of the emitters reporting data that differed from pollutant to pollutant in comparison to the estimated data. Out of those, 30% of the emitters reported smaller and 8% reported greater emissions than estimated emissions with 7% of the emitters not submitting any emissions data. The analysis also includes calculation of the emissions scope limit within 95% confidence interval. According to these results, it can be concluded that only 15% of the emitters have emission levels that fall within the scope limits, 47% of the emitters reported data of which two data fall within the scope limits, 28% of the emitters reported data of which only one data fall within the scope limits, 3% of the emitters had data that didn’t fall within the scope limits, and 7% did not reported any emission data. The results of the analysis can be summarized as: 1) operators in facilities do not know how to calculate emissions from their sources, and 2) the application of global emission factors can lead to considerable errors. The main reasons for significant deviation are different fuel quality, type of technical units and human error, thus national emission factors should be developed