Ammonia (NH₃) volatilization was measured following soil-surface application of poultry manures obtained from contrasting handling and storage systems using wind tunnels and semi-open chambers. Our objectives were 1) to quantify NH₃ losses from poultry manures typical of Eastern Canada poultry farms, 2) to relate emissions to manure properties, and 3) to compare volatilization estimates using wind tunnels and semi-open chambers. Broadcasting poultry manures resulted in NH3 losses accounting for 13 to 35% of total manure N. Compared to wind tunnels, the use of semi-open chambers for measuring volatilization losses results in two types of estimation biases. The instantaneous NH₃ emission rates were strongly underestimated (70%) as a result of the slow diffusion-based gas exchange in the semi-open chamber headspace. This underestimation could be at least partially corrected by mechanical air mixing using fans. We also found that the more humid conditions inside the semi-open chambers enhanced the mineralization rate of manure nitrogen, thereby increasing the NH₄-N concentration and the NH₃ emissions. Therefore, the differences between the wind tunnel and chambers estimates varied depending on the manure properties (organic N content) and the length of the measurement period; stressing the importance of including contrasting manures and describing temporal volatilization dynamics in future studies comparing chamber methods. We conclude that land application of poultry manures can be a large source of NH₃, that cumulated volatilization losses are proportional to their ammonium-N content and that semi-open chambers provide strongly biased estimates of NH₃ losses.