Thermal-optical analysis is a conventional method for determining the carbonaceous aerosol fraction and for classifying it into organic carbon, OC, and elemental carbon, EC. Unfortunately, the different thermal evolution protocols in use can result in a wide elemental carbon-to-total carbon variation by up to a factor of five. In Europe, there is currently no standard procedure for determining the carbonaceous aerosol fraction which implies that data from different laboratories at various sites are of unknown accuracy and cannot be considered comparable. In the framework of the EU-project EUSAAR (European Supersites for Atmospheric Aerosol Research), a comprehensive study has been carried out to identify the causes of differences in the EC measured using different thermal evolution protocols; thereby the major positive and negative biases affecting thermal-optical analysis have been isolated and minimised to define an optimised protocol suitable for European aerosols. Our approach to improve the accuracy of the discrimination between OC and EC was essentially based on four goals. Firstly, charring corrections rely on faulty assumptions – e.g. pyrolytic carbon is considered to evolve completely before native EC throughout the analysis –, thus we have reduced pyrolysis to a minimum by favoring volatilisation of OC. Secondly, we have minimised the potential negative bias in EC determination due to early evolution of light absorbing carbon species at higher temperatures in the He-mode, including both native EC and combinations of native EC and pyrolytic carbon potentially with different specific attenuation cross section values. Thirdly, we have minimised the potential positive bias in EC determination resulting from the incomplete evolution of OC during the He-mode which then evolves during the He/O₂-mode, potentially after the split point. Finally, we have minimised the uncertainty due to the position of the OC/EC split point on the FID response profile by introducing multiple desorption steps in the He/O₂-mode. Based on different types of carbonaceous PM encountered across Europe, we have defined an optimised thermal evolution protocol, the EUSAAR_2 protocol, as follows: step 1 in He, 200 °C for 120 s; step 2 in He 300 °C for 150 s; step 3 in He 450 °C for 180 s; step 4 in He 650 °C for 180 s. For steps 1–4 in He/O₂, the conditions are 500 °C for 120 s, 550 °C for 120 s, 700 ° C for 70 s, and 850 °C for 80 s, respectively.