The technique for the analysis of soil air for small quantities of hydrocarbons is reviewed. Various methods of expressing results and the basis for the calculation of each are given. Field data on expansion ratios are presented to show that there exists in the soil air a gas, other than the saturated hydrocarbons, which is not removed by the screening agents and is retained in the condensation trap. The reasons for believing that this gas is nitrous oxide are given, together with plausible sources of this gas in the soil air. Laboratory data are presented on the analysis of pure hydrocarbons in the presence and in the absence of this gas. Calculations made by each of the methods show that when results are based on after‐burning readings, a true measure of the hydrocarbon content is evolved, regardless of the presence of nitrous oxide.