Introduction A key barrier when restoring arid ex-grazing land is the presence of artificial watering points (AWPs). Even after the cessation of domestic livestock grazing, AWPs support high numbers of wild herbivores and sustain high grazing pressure, suppressing plant species establishment. One approach to reducing total grazing pressure is to decommission AWPs; however, it is unclear how this impacts vegetation and groundcover recovery, or the efficacy of different closure methods.Objectives This study aimed to assess the impact of AWP closure methods on vegetation and groundcover in relation to distance from AWP, time, and rainfall.Methods In this 9-year study on ex-grazing land now managed for conservation, AWPs were experimentally closed using three closure methods, plus AWPs left "open" as controls. Vegetation and groundcover were measured at increasing distances from AWPs each year.Results Two years of extremely high rainfall resulted in a reversal of the piosphere effect at all AWPs regardless of closure treatment. Bare ground consistently decreased over time, and native plant cover in dunes increased. The treatment effects close to AWPs were confounded by disturbance effects, particularly at AWPs where earthmoving redistributed the soil profile.Conclusions Our results reflect the broader recovery of an ecosystem on a restoration trajectory, where outcomes have manifested after a high rainfall period. Uncertain results from this work and other AWP closure studies reflect the challenges in designing AWP closure studies at ideal spatial and temporal scales.