The response of soil respiration to the increasing nitrogen deposition plays an important role in predicting carbon cycle of terrestrial ecosystems. To date, however, few studies have been addressed to soil respiration changes in shrub ecosystem compared with forest and grassland ecological system. In this study, we subjected Sibiraea angustata shrub ecosystem, the main type in the eastern margin of Qinghai-Tibetan Plateau, to four nitrogen addition levels practiced within 5 m×5 m plot: N₂(2 g N m^-2 a^-1), N₅(5 g N m^-2 a^-1), N₁₀(10 g N m^-2 a^-1) and N₀(control, add the same volume of water without nitrogen in other treatments), respectively. Ditching method was used to separate the different components of soil respiration. Plant root respiration was defined as autotrophic respiration, while soil microbial and animal respirations were defined as heterotrophic respiration. Changes of soil respiration to the short-time N addition were measured during the growing season, with the soil moisture at 5 cm and 10 cm depth measured at the same time. The results showed: (1) Soil respiration (total, heterotrophic and autotrophic respiration) showed apparent seasonal variation during the experimental period (2012.5-2012.10). During the growth season, nitrogen addition had no significant effects on total and heterotrophic respiration (P > 0.05), while it remarkably inhibited the autotrophic respiration (P < 0.05). It turned out that the soil respiration rates under N₂ and N₁₀ treatment was lower than N₀ treatment by 0.31 and 0.23 μmol m^-2 s^-1, respectively. Meanwhile, the soil respiration rates under N₅ and N₀ treatment exhibited no obvious difference. (2) Soil respiration (the total, heterotrophic and autotrophic respiration) showed apparent diurnal variation. Nitrogen addition had significant effects on the total, heterotrophic and autotrophic respiration (P < 0.001) in the whole day. Nitrogen addition remarkably promoted the heterotrophic respiration, with soil respiration rates under N₂, N₅ and N₁₀ treatment were higher than N₀ treatment by 0.81, 0.84 and 1.50 μmol m^-2 s^-1, while it remarkably inhibited the autotrophic respiration, indicating that the soil respiration rates under N₂, N₅ and N₁₀ treatment was lower than N₀ by 0.46, 0.99 and 0.48 μmol m^-2 s^-1, respectively. (3) On the whole, nitrogen addition significantly inhibited soil carbon dioxide emissions of autotrophic respiration over the experimental period (P < 0.05), showing that the carbon dioxide emissions under N₂ and N₁₀ treatment was lower than N₀ treatment by 11.75, 8.50 kg hm^-2 d^-1. Nitrogen addition had a significant effect on the soil carbon dioxide emissions of autotrophic respiration in May and July, but the influence of different nitrogen concentration was not regular, while it promoted the soil carbon dioxide emissions of heterotrophic respiration only in May. (4) The total and heterotrophic respiration showed significant exponential correlation with the soil temperature at 5 cm depth (P < 0.001), and the nitrogen addition strongly affected the sensitivity of soil respiration to temperature change. Besides, the soil respiration showed a weaker correlation with the soil moisture compared with temperature, which demonstrated that the soil moisture might not be the limiting factor to the soil respiration in the experimental region. The soil respiration is the main link in carbon cycle of terrestrial ecosystems, and the increasing nitrogen deposition would affect soil respiration deeply through a series of biochemical processes. Therefore, the underlying mechanism of how the components of soil respiration response to N deposition still needs further stduy.