It is commonly believed that ecosystem responses to global climate change are nonlinear. However, nonlinear response patterns are not well characterized on ecosystem carbon and water processes. We used a terrestrial ecosystem (TECO) model to examine nonlinear patterns in responses of a grassland ecosystem in central USA to 100-year gradual changes in temperature, CO₂, and precipitation individually or in combination. Variables examined in this study are net primary productivity (NPP), Rh (heterotrophic respiration), net ecosystem exchange of CO₂ (NEE), runoff, and evapotranspiration (ET). Our modeling results show that nonlinear trajectories are parabolic, asymptotic, and threshold-like in response to temperature, CO₂, and precipitation anomalies, respectively, for NPP, NEE, and Rh. Runoff and ET exhibit threshold-like pattern in response to both temperature and precipitation anomalies, while they are insensitive in response to CO₂ changes. The combined two- or three-factor changes in temperature, CO₂, and precipitation considerably influence ecosystem nonlinear responses by either changing patterns or shifting points of abrupt changes. Our modeling results suggest that nonlinear response patterns to global climate change are diverse and were interactively affected by other global change factors on ecosystem carbon and water processes.