Surface water availability in drylands has changed with the introduction of artificial water points. Despite known ecological impacts, detailed mapping of this change has not occurred in most drylands. We aimed to quantify the extent and distribution of changes in water availability. We tested whether water availability increased more in pastorally productive areas than less fertile areas, and whether remaining water remote areas are restricted to low productivity landscapes. Our new spatially-explicit method mapped access to water at fine spatial scale, weighting locations by their distance to water and the permanence of those water sources. We demonstrated our method in a study area of over 700,000 km2 in Queensland, Australia, with our mapping showing large changes in water availability since pastoral development. Less than 5% of the study area is now more than 10 km from water, compared with almost 60% previously. Few refuges for grazing-vulnerable communities remain. Even low fertility landscapes showed marked increases in water availability. This has conservation implications for managing production landscapes. Our approach can be applied in any dryland landscapes that have experienced changes in water availability, and can help guide actions such as removing artificial water points to recreate ecological refuges.