【Objective】 Nitrogen (N) release characteristics of controlled release urea (CRU) is affected by properties of the fertilizer per se and environmental conditions, such as climate, soil temperature and soil moisture etc. The static water extraction method is widely used for quality control of CRU. However, as a matter of fact, CRU is mainly applied to farm fields, where soil moisture is the major factor affecting N release from CRU. Soil moisture can be characterized by soil water suction, which is closely related to the process of nutrient transmembrane movement in CRU. So it is essential to explore threshold value of the soil water suction that controls N release rate of CRU so as to provide a scientific basis for proper application of CRU in arid regions. 【Method】 Static water extraction and soil culture experiments were carried out in this study to investigate N release characteristics of CRU in water and soil conditions, separately, under constant temperature of 25℃. The static water extraction experiment of CRU was performed following the industry standard for controlled release fertilizer (HG/T 4215-2011). In line with the soil water characteristic curves of the tested loamy sand soil (fluvo-aquic soil, Ochri-Aquic Cambosols), five levels (0, 30, 75, 260 and 525 kPa) of soil water suctions were designed for the soil culture experiment, and air-dried soil was used as control. And, vapors from the soils of the six treatments were measured with the simulated incubation method in the desiccator. 【Result】 Results show that N release of the CRU in static water was quite similar to that in the soil 0 kPa in soil water suction. Cumulative release rate in the first seven days increased very slowly, sped up from the 7th to the 14th day, and then leveled off after the 21st day. The CRU released 80% of its N in 55.9 days and 63.6 days, respectively, in water extraction and soil incubation. N release varied similarly in characteristics, peaked in the period from the 20th to the 30th day, and reached 18.3%, 22.2%, and 15.3%, respectively, in the treatments, 75 kPa, 30 kPa, and 0 kPa in soil water suctions, in 20 days. As the air in soil pores was > 95% in relative humidity in all the tested soils, the soils were all the same in soil water vapor saturation and N release period, which indicates that in this case, excluding the impacts of flowing soil water and nutrient diffusion, soil moisture is no longer the major factor affecting N release from CRU. N release in the treatments, 525 kPa and 260 kPa in soil water suction and 84% and 91% in relative humidity of pore air, lasted for 416.4 d and 120.0 d, respectively, which was 146.8% and 59.1%, longer relative to that in the treatment saturated with water (63.6 days) and 20% longer relative to that set in the industry standards for controlled-release fertilizer, indicating that N release rate and N releasing period of the CRU is affected by too high soil water suction. 【Conclusion】 All the findings of this study demonstrate that the differences between the six treatments in N release characteristic of CRU are striking, when impacts of water flow and N diffusion are not taken into account; Soil moisture is no longer the main factor affecting N release of CRU when the soil is less than 75 kPa in soil water suction; In soils with soil water suction being higher than 260 kPa, N release rate and duration of the CRU are restrained by soil moisture. Soil water suction is directly related to saturation of pore air in humidity. The influence of soil water suction on release characteristic of CRU is affected through the impact of vapor in the soil on fertilizer granules.