In this paper we present an overview of the present knowledge relating to methods that avoid interference of N2O on δ13C and δ18O measurements of CO2. The main focus of research to date has been on atmospheric samples. However, N2O is predominantly generated by soil processes. Isotope analyses related to soil trace gas emissions are often performed with continuous flow isotope ratio mass spectrometers, which do not necessarily have the high precision needed for atmospheric research. However, it was shown by using laboratory and field samples that a correction to obtain reliable δ13C and δ18O values is also required for a commercial continuous flow isotope ratio mass spectrometer. The capillary gas chromatography column of the original equipment was changed to a packed Porapak Q column. This adaptation resulted in an improved accuracy and precision of δ13C (standard deviationGhent: from 0.2 to 0.08‰; standard deviationLincoln: from 0.2 to 0.13‰) of CO2 for N2O/CO2 ratios up to 0.1. For δ18O there was an improvement for the standard deviation measured at Ghent University (0.13 to 0.08‰) but not for the measurements at Lincoln University (0.08 to 0.23‰). The benefits of using the packed Porapak Q column compared with the theoretical correction method meant that samples were not limited to small N2O concentrations, they did not require an extra N2O concentration measurement, and measurements were independent of the variable isotopic composition of N2O from soil. Copyright © 2005 John Wiley & Sons, Ltd.