AbstractAmmonia volatilization from agricultural land has a negative impact on the atmospheric, aqueous, and land environments. However, this process is still not fully understood. In this study, a 2‐year field experiment is conducted in the Taihu Lake region of China to monitor ammonia volatilization from a paddy soil. The ammonia volatilization losses are in the range of 43.4 to 53.6 kg N ha−1 in the year 2013 and 25.9 to 35.4 kg N ha−1 in the year 2014, which account for 19.7% to 19.9% and 11.8% to 13.1% of the applied nitrogen with two different N treatments (N1: 220 kg N ha−1, N2: 270 kg N ha−1). The ammonia volatilization loss is highest during the basal fertilizer application, followed by the tillering and the ear‐differentiation stages. Then, two empirical models and one mechanistic model are used to simulate the ammonia volatilization process. These models are calibrated and validated with field experimental data. The results show that none of the three models accurately predict the ammonia volatilized from the rice field (R2 < 0.40), however, among the three models evaluated, the mechanistic model is more effective in simulating the ammonia volatilization process (R2 = 0.38).