This work aims at gaining information on the effects of different thermal treatments on radiocarbon measurements of organic (OC) and elemental (EC) carbon fractions in the atmospheric aerosol. Improvements to the traditional approaches for the determination of the fraction of modern carbon of OC and EC- f(m)(OC) and f(m)(EC)- are proposed. As for f(m)(EC) determination, we propose to add a He step at high temperature to the standard oxygen treatment Our tests demonstrate that the addition of a high temperature He step (final choice: 750 degrees C) to the oxygen treatment is effective in removing the refractory OC. As for f(m)(OC) determination, we propose to quantify it by measuring either the fraction of modern carbon of TC and EC or the fraction of modern carbon of water soluble (f(m)(WSOC)) and water insoluble (f(m)(WINSOC)) organic carbon to limit the influence of possible pyrolysis on the direct determination of f(m)(OC). Tests on the equivalence of the approaches have shown good agreement between them. Our tests were carried out on samples collected in a heavily polluted area (Milan, Italy) during wintertime. f(m)(OC), f(m)(EC), and f(m)(TC) values obtained in our tests were also used to attempt a preliminary source apportionment in the area using C-14 measurements. Wintertime EC resulted to be mainly fossil (84%), whereas OC was dominated by modern contribution (63%). Two approaches were tested for the evaluation of the wood burning contribution to OC and a good agreement was found. Wood burning primary contribution accounted for about 18% of OC in Milan during wintertime. Secondary OC from biomass burning and the contribution from other urban sources were tentatively identified following literature approaches, with the aim of evaluating the biogenic contribution to OC in the area, which was estimated to be about 18%. (C) 2012 Elsevier Ltd. All rights reserved.