Groundwater and surface water-fed systems act as biodiversity hotspots and ecological refuges and evolutionary refugia in arid regions. Groundwater-dominated systems are sustained by underground aquifers that are recharged by rain that has fallen in the distant past, while surface water-dominated systems are fed by recent local rain or floods. Some waterbodies are fed by a mixture of these sources. Perennial, groundwater-dominated systems will act as refuges and refugia under future rainfall declines associated with global warming. We sought to identify climate refugia, based on groundwater dominance, by using isotope hydrology to characterise water samples collected by citizen scientists across arid central Australia. There is a linear relationship between hydrogen isotopes (2H/1H, δ2H) and oxygen isotopes (18O/16O, δ18O) in rainfall. This relationship is known as the meteoric water line (MWL). By comparing our samples with the Australian MWL, and developing a local evaporation line, we were able to test the hypotheses that groundwater-dominated systems will follow the Australian MWL while temporary systems follow the local evaporation line, and, accordingly, distinguish between groundwater and surface water-dominated systems. The isotopic composition of samples collected over a 36-month period was determined using isotope ratio infrared spectrometry. The electrical conductivity of each sample was recorded to determine where freshwater is available for biota within this arid region. Over 240 water samples were collected from 62 waterbodies and seven bores (groundwater wells) spanning an area of more than 250,000 km2. Approximately 75% of the samples were collected by citizen scientists and 25% by research scientists. Twenty groundwater-dominated waterbodies, characterised by a small range of δ2H and δ18O values (c. −55 to −20‰ and c. −9 to −3‰, respectively) clustered around the long-term mean composition of rainfall (δ2H = −37.5‰, δ18O = −6.4‰), were identified as future evolutionary refugia. These sites are likely to contain water through the most severe of droughts and will be critically important for the persistence of water-dependent species. Based on their isotopic composition, we identified 45 waterbodies (rockholes/waterholes) as temporary or ephemeral (δ2H c. −40 to −100‰ and δ18O c. −4 to +25‰), that is, with no evidence of groundwater inflow. These, together with waterbodies supported by a mix of groundwater and surface water, can act as stepping stones and form part of the aquatic mosaic that is critical to supporting species in arid regions. Over two-thirds of the waterholes sampled were very fresh (electrical conductivity <0.8 mS/cm), indicating that they provide the freshwater needed to support much of the regional aquatic and terrestrial fauna. All evolutionary refugia are located within protected areas (i.e. national parks or Indigenous Protected Areas), but some are subject to the impacts of feral animal species and invasive plants. Our findings indicate where control programmes and restoration actions can be prioritised to support biodiversity conservation and climate change adaptation. Our approach, combining citizen science and isotope hydrology, can be used to identify future refugia in other remote and arid regions where water scarcity is likely to increase under global climate change.