Methane (CH4) emissions from a herd of 58 grazing cattle were determined in a field experiment using paddock-scale (micrometeorological) methods. The emissions were also measured daily from each animal, using the sulfur hexafluoride tracer method. The paddock-scale methods exploit how the gas, once emitted from the cattle, is transported and dispersed by the wind. Hence, the emission rate may be calculated from measurements of windspeed, wind direction and turbulence statistics, as well as CH4 concentration upwind and downwind of the herd. The paddock-scale methods include a mass-budget approach, flux-gradient method and gas dispersion model. Accuracy can be assessed in unprecedented detail because the animal-scale (reference) method included all individuals in the herd, and the measurement site was ideal for micrometeorological methods (flat, usually windy and free of obstructions that alter the turbulent airflow). The cattle were hand-reared steers of average weight 325 ± 20 kg. Based on the animal-scale method, the average CH4 emission rate over 9 days was 161 ± 20 g/steer.day. The gas dispersion model, when utilising vertical concentration profiles, yielded on average 27% greater emissions. The other paddock-scale methods agreed with the animal-scale method, provided the cattle were at least 22 m away from the location of the downwind concentration measurements. When the cattle were allowed to graze as closely as 5 m from the instruments, the paddock-scale methods gave greater emissions than the animal-scale method; reasons for this are discussed.