ABSTRACT Monitoring salt pans is important for agricultural management in arid/semi-arid regions because salt pans can negatively affect human life, wildlife, ecology, and cause accelerated desertification, cropland loss, and economic downturn. Spectral salt pan indices based on remotely-sensed data from previous studies vary by location and may not be readily applicable to another location due to the spatial heterogeneity of salt components across the continental surface. Using Landsat-8 imagery and climate datasets, this study aims to develop a mapping scheme which effectively delineates the maximum salt pan/playa extents under various spectral conditions for different geographic locations. Calibrated by a spectral training pool sampled from eight major salt pan/playa sites in the world, this mapping scheme first implements a conservative spectral index to highlight potential salt-covered regions, from which salt pans/playas are further extracted by a support vector machine (SVM). Quality assurance/quality control (QA/QC) with assistance of auxiliary datasets and high-resolution images was performed to reduce commission errors and refine the results. The mapping scheme was applied to the endorheic regions of the western United States with a total area of approximately 700,000 km2, of which 1.6% was mapped as salt pans. Quality assessments show reasonable reliability of the mapping result for most regions, with the producer’s accuracy, user’s accuracy, and kappa coefficient exceeding 90%. The mapped spatial details reveal that salt pans are distributed fairly evenly across the arid western United States, with areas positively correlated to the sizes of the endorheic basins. Given these results, we foresee potential applicability of the developed scheme to map salt pans/playas across Earth’s continental surface. Our spectral training pool and produced salt pan/playa extents are available through https://doi.pangaea.de/10.1594/PANGAEA.913152.