Abstract:[Objective] Ammonia volatilization from solar greenhouses driven by excessive nitrogen input has received widespread attention in recent years. However, the ammonia volatilization emission of greenhouse system is mainly determined by soil surface ammonia volatilization. Greenhouses are semi-closed structures, as a result, a part of NH3 emitted from soil can be absorbed again by the plant canopy or dissolved in the greenhouse membrane water and returned to the soil. The volatilization of ammonia from soil surface can not accurately reflect the amount of ammonia discharged into the atmosphere by solar greenhouse. Therefore, ammonia volatilization from soil surface is difficult to accurately reflect the amount of ammonia emitted into the atmosphere by solar greenhouse.[Method] Experiment comprised four treatments:(i) no nitrogen input with conventional irrigation (N0+FI), (ii) conventional nitrogen input with conventional irrigation (FT+FI), (iii) optimum nitrogen input with conventional irrigation (OPT+FI), and (iv) optimum nitrogen input with optimum irrigation (OPT+OI). Ammonia volatilization losses were measured in three vegetables seasons (tomato-watermelon-tomato) using intermittent closed chamber ventilation method. The air volume mask (Kanomax 6570) was used to measure the gas flow at the vent, and the ammonia concentration at the vent was continuously monitored by pumping method. In this way, the loss rate and amount of ammonia volatilization were measured from the whole greenhouse, and losses were compared with the soil surface.[Result] The results showed that the ammonia volatilization rate peaked on the same day after greenhouse fertilization, and there was no significant difference between fertilized and unfertilized (control) after 7 days. During the three planting seasons, the ammonia volatilization under different nitrogen treatments were:2.82-4.97 kg·hm–2, 6.59-9.97 kg·hm–2and 15.77-21.83·kg hm–2, respectively, and the corresponding ammonia volatilization emission factors were 0.64%-1.50%, 3.11%-4.21% and 2.59%-3.90%, respectively. The trend of ammonia volatilization rate of the whole shed was basically consistent with that of the soil surface. The ammonia volatilization rate of the whole shed was N 2.22 kg·hm–2 in the second quarter and N 2.92 kg·hm–2 in the third quarter, which accounted for 13.38%-33.69% of the ammonia volatilization from the soil surface, and the ammonia volatilization coefficient was only 0.46%-1.48%, which was much lower than the ammonia volatilization from the soil surface.[Conclusion] Thus, it is concluded that the ammonia volatilization from solar greenhouse will be overestimated when only taking the soil surface ammonia volatilization into consideration. Measuring ammonia volatilization based on the whole solar greenhouse system is thus recommended in future studies.