The intensity and frequency of precipitation events, especially extreme precipitation events (EPE), are increasingly recognized for their importance in regulating ecosystem functions (e.g., C fluxes). Here we assessed the effects of precipitation events on ecosystem carbon exchange following the precipitation events. Naturally occurring precipitation events and CO₂exchange were measured by applying an eddy covariance technique and meteorological tower between 2008 and 2010 in a desert grassland in Inner Mongolia, China. The results indicate that the accumulative net ecosystem exchange (NEE) of carbon was -11.08 gC·m^-2, -15.62 gC·m^-2 and 26.68 gC m^-2 in 2008, 2009 and 2010, respectively. The magnitude and timing of precipitation controlled the variation in accumulative NEE. Summer and autumn precipitation produced stronger impacts than those in the spring. The desert grassland appeared more sensitive to precipitation than other grassland ecosystems, with a threshold of ～1–2 mm precipitation. During the 3-year study period, EPEs significantly enhanced the carbon sink strength due to discrepant sensitivity of gross ecosystem productivity (GEP) and ecosystem respiration (R_eco) to precipitation events. NEE response time to EPE (T_L) lasted 5–17 days. Total accumulative carbon was - 40.32 gC m^-2, which was 91% more than the 3-year average of NEE during T_L and 33% more than the growing season average. GEP and R_eco were disproportionally promoted by 52% and 37% with EPEs during T_L. A nonlinear equation (NEE = α + βte^-γt) was developed with high confidence (R²= 0.84) to model the changes of NEE following the precipitation event. The sensitivity response of carbon exchange processes to larger precipitation events suggests that the carbon sink strength of the Inner Mongolian desert grassland will be elevated with the increase in the frequency and intensity of extreme precipitation. This study provides updated evidence regarding how precipitation events alter ecosystem carbon sequestration and helps mitigate the uncertainty when quantifying the carbon budget of a semi-arid ecosystem with a different future climate. Further studies are crucial to investigate the long-term responses of carbon exchange processes to variations in precipitation patterns and characterize the responses of biological components of the carbon balance following precipitation events.