The enigma of soil nitrogen balance sheets has been discussed for over 40 years. Many reasons have been considered for the incomplete recovery of 15N applied to soils, including sampling uncertainty, gaseous N losses from plants, and entrapment of soil gases. The entrapment of soil gases has been well documented for rice paddy and marshy soils but little or no work appears to have been done to determine entrapment in drained pasture soils. In this study 15N-labelled nitrate was applied to a soil core in a gas-tight glovebox. Water was applied, inducing drainage, which was immediately collected. Dinitrogen and N2O were determined in the flux through the soil surface, and in the gases released into the glovebox as a result of irrigation or physical destruction of the core. Other components of the N balance were also measured, including soil inorganic-N and organic-N. Quantitative recovery of the applied 15N was achieved when the experiment was terminated 484 h after the 15N-labelled material was applied. Nearly 23% of the 15N was recovered in the glovebox atmosphere as N2 and N2O due to diffusion from the base of the soil core, convective flow after irrigation, and destructive soil sampling. This 15N would normally be unaccounted for using the sampling methodology typically employed in 15N recovery experiments.