Abstract. The aim of the study was to measure the carbon 14 activity at natural level in air samples using classical methods of radiochemistry and beta counting. Three different methods have been tested in order to minimise the detection limit. In the three methods, the first step consists in trapping the atmospheric carbon 14 into NaOH (1N) using a bubbling chamber. The atmospheric carbon dioxide reacts with NaOH to form Na 2 CO 3 . In the first method the Na 2 CO 3 solution is mixed with a liquid scintillate and is directly analysed by liquid scintillation counting (LSC). The detection limit is approximately 0.3 Bq.m -3 of air samples. The second method consists in evaporating the carbonate solution and then counting the solid residue with a proportional gas circulation counter. The detection limit obtained is equivalent to the first method (0.36 Bq.m -3 of air samples). In the third method, Na 2 CO 3 is precipitated into CaCO 3 in presence of CaCl 2 . CaCO 3 is then analysed by LSC. This method appears to be the most appropriate, the detection limit is 0.04 Bq.m -3 of air samples.

No abstract