High-quality emission data are of utmost importance for the reliable outputs of air quality models. AQ models require speciated hourly emissions, whereas emission inventories are constructed on annual basis and include only main pollutants. For the time disaggregation simple and constant monthly, daily, and hourly factors are usually used. Likewise, chemical speciations are usually based on outputs of very few and sometimes quite old reports and do not necessarily correspond to the situation in the country of interest. For the emission processing itself many scientific groups use SMOKE pre-processor, due its availability and since it is fitted for the WRF and CMAQ models. Nevertheless, its application for non-US countries can be quite complicated due to different data and metadata structure. Our project focusses on two aims: 1) The derivation of the time and speciation profiles specific for the Czech Republic. For this purpose, we analyse continuous emission measurements from the large stationary sources, temperature and behaviour dependency of the household heating emission, and the data available in the extensive Czech emission database REZZO. For the transport we managed to access data from the Electronic Control Units (ECU) of the vehicular fleet maintained by the SHERLOG company. These unique data give us information on engine state, fuel consumption, and driving styles from the real traffic flow. 2) The construction of a flexible emission processor for AQ models based on the freely available software.