The tracer-ratio technique has been used to measure methane emissions from dairy cattle housed in barns and on feedlots and for ammonia emissions from livestock buildings. This study evaluates the application and reliability of the tracer gas ratio method for measuring the emission rates of methane and nitrous oxide from naturally ventilated swine hoop structures. A tracer gas dispersion system was built for dispersing the tracer gas SF6. The sampling grid system included three planes that intersected the plume, with the first perpendicular to the plume at the effluent face of the hoop in the ZX plane, the second downwind in the vertical YZ plane of the plume, and the third downwind in the XY plane 0.76 m above the sod surface. Air samples were taken manually and analyzed for CH4, N2O and SF6 using a lab-based gas chromatograph. Variation in wind speed resulted in fluctuating downwind concentrations in the plumes. The SF6 tracer gas plume did not consistently coincide with the plumes of the other gases at all downwind points in the ZY plane, indicating non-uniform sources and inadequate mixing of the gases. These differences in the locations of the plumes resulted in considerable variability of calculated emission rates. With increasing distance from the hoop more uniform mixing of the gases was observed. At 3.04m from the hoop there was less variability in the calculated emission rates, presumably indicating that these rates were more representative of actual building emissions. The presence of pigs in the hoop resulted in not only higher emission rates of CH4, and N2O but also greater variability in calculated emission rates. Variation in the height of the dispersion rack did not significantly influence the measured gas emissions.