Vegetation pattern morphology is suggested as one indicator of system closeness to desertification. Using pattern morphology as an indicator requires understanding the timescales at which patterned vegetation systems respond to drought. Modeling these timescales requires accounting for rainfall intermittency and all the pathways controlling vegetation-precipitation feedbacks. In this paper, a model of rainfall initiation and intensity based on the dynamics of a single-column atmospheric boundary layer is coupled to a patterned vegetation model. The coupled climate-vegetation model, parameterized to represent a typical vegetation morphology in southwestern Niger, is used to investigate the timescales of desertification due to shifts in the total annual rainfall regime, and the effect of precipitation feedbacks on these timescales. Depending on the exact rainfall history, biomass and spatial morphology may not respond monotonically to a decrease in rainfall. The model results suggest changes in pattern morphology responding to shifts in annual rainfall require at least 4–5 years. Feedbacks acting through vegetation's influence on surface albedo and, to a lesser extent, surface evapotranspiration act to speed up the vegetation response to drought. Although the overall local-scale vegetation-precipitation feedback is positive, individual storm events may exhibit negative feedbacks, in which rainfall only occurs for low vegetation cover, depending on the free atmospheric conditions. Vegetation-precipitation feedbacks are sufficiently important to speed up changes in vegetation patterns, even in marginal drylands with low biomass levels.