One significant feature of the North American Monsoon (NAM) is its sudden onset and the accompanying rapid greenness of vegetation cover. The spatial and temporal vegetation variability can modify the surface heat and moisture fluxes and, in turn, influence atmospheric process, such as mesoscale circulation and the formation of clouds and precipitation systems. Releastic representation of the vegetation response (i.e., the change in live biomass) to atmospheric and hydrological forcings is currently lacking in the land-surface parameterizations used in NAM study. In this paper, we first examine if the CENTURY biogeochemistry model is able to realistically simulate vegetation growth for the monsoon region. Observed atmospheric forcings dervied from the first-order summary of the day (SOD) station data were used to drive CENTURY (DayCENT) at a daily time step to simulate vegetation growth over southwestern U.S. and Mexico from 1982 to 1998. Six perturbation runs were performed, with Tmax and Tmin increased and decreased two degrees Centigrade, and the precipitation increased or decreased by 25% from their original values. These results are then compared with equivalent simulations for Central U.S. made in a previous study (Lu et al, 2001) for the same time period to investigate regional differences in vegetation response to atmospheric forcings. Meanwhile, the DayCENT-simulated leaf area index (LAI) were validated against Pathfinder AVHRR NDVI-derived LAI. A comparison between the CENTURY and RAMS hydrological cycles was also performed to investigate consistency within the coupled system.