Obtaining representative gaseous emission factors from livestock production requires measurement methods adapted to a high number and a great diversity of livestock systems, including naturally ventilated buildings. The objective of this study is to propose such a method for cattle houses, based on livestock-related data (e.g., feeds, production, effluents) and intermittent measurements of gas concentrations (H2O, CO2, NH3, CH4, and N2O), temperature, and relative humidity. These data were used to estimate gas-concentration gradients and mass-balance deficits of C, N, and H2O). Emissions were calculated using the ratios of gas-concentration gradients. For quality control, the results were validated with water-budget observations and simulation model predictions of CH4 and CO2 emissions from cattle. During summer 2007, measurements were performed in seven types of barns commonly found in France. For buildings with slurry evacuated twice a day, good agreement was found between CO2 and CH4 emissions estimated with the simplified method and those predicted from models of cattle emission. For these buildings the observed emissions were homogeneous. For buildings with deep litter, observed emissions of CH4 and CO2 were higher than the predicted emissions. The difference indicates that a part of those gases was emitted by manure. Additional data and models should be used to improve this method for deep-litter systems. Data analysis continues to evaluate the method during winter conditions and in poultry or swine farms.