The effects of cattle farming systems and individual management practices on GHG emissions have been studied by applying a screening method on eight European countries totalizing 60 dairy farms and four seasonal measurements. This method is based on indoor and outdoor CO2, CH4, and N2O concentration measurements and on a questionnaire developed to estimate the carbon mass balance at the building scale. The results of the questionnaire were also used to characterize the farm diversity. The dairy population in the barn where the air samplings were done ranged from 8 to 979 cows. The feeding quantity was between 14 and 31 kg DM/animal/day and the milk production between 18 and 50 kg milk/animal/day. Most of them had concrete or slatted floors (40% and 35% respectively) and one fourth were separated urine and feces. Minimum difference to deduce emission estimates were 9 mg m−3 for CO2 and 0 mg m−3 for CH4, NH3 and N2O. We also considered a minimum threshold of 5 for the ratio (C − CO2, in−C − CO2, out) · (C − CH 4, in − C − CH4, out) −1 because of enteric emissions in dairy houses. Almost 90% of observations were above these thresholds except for N2O. This international survey conducted with the same measurement method helped identifying the main factors influencing GHG emissions as well as the existing low-emitting systems: the calculated emissions based on concentration measurements showed that CH4 emissions from manure could double those from the enteric fermentation of the animals; the detection level of the N2O emissions was below 1 g N-N2O/animal/day; in some farms, the emission results showed very few seasonal variations over the year while, for other ones, seasonal variations were well observed; the results showed well detectable GHG emission reductions related to changes in farm practices.