AbstractWater management affects NH3 emission in complex ways due to its influence on nitrogen transformation and movement. This study aimed to determine the effect of irrigation timing on NH3 emission from urea at different initial moistures. A pot experiment was conducted based on three typical soils (Fluvisols, Cambisols, and Luvisols) in the North China Plain. Each soil was treated to three levels of initial moisture (5%, 10%, and 15% w/w) and irrigated with 40 mm water 0, 6, and 12 hr after applying urea. The results showed that NH3 loss ranged from 0.2 to 14.2 kg N ha−1, with the greatest loss in Fluvisols, followed by Cambisols, and the lowest in Luvisols. The loss of NH3 increased with initial moisture for all three soils. But there was no significant effect of irrigation time on NH3 emission, although the contents of NH4+ in surface soil were higher for delayed irrigation than for immediate irrigation. Additionally, the NO3− content was increased in Luvisols with delayed irrigation at a high initial water content, which may have promoted leaching. Therefore, delaying irrigation by up to 12 hr did not cause a significant increase in NH3 emission. However, fertilization and irrigation should be carried out with lower moisture, especially for Luvisols to prevent ammonia emission or nitrate leaching.