Defining the optimal placement of areas for biodiversity conservation in developing nations remains a significant challenge. Our best methods for spatially targeting potential locations for biodiversity conservation rely heavily on extensive georeferenced species observation data which is often incomplete or lacking in developing nations. One possible solution is the use of surrogates that enable site assessments of potential biodiversity values which use either indicator taxa or abiotic variables, or both. Among the plethora of abiotic variables, soil carbon has previously been identified as a potentially powerful predictor for threatened biodiversity, but this has not yet been confirmed with direct observational data. Here we assess the potential value of soil carbon for spatial prediction of threatened species using direct measurements as well as a wide range of GIS derived abiotic values as surrogates for threatened plant species in the PEBANPA network of permanent plots in Southern Patagonia. We find that soil carbon significantly improves the performance of a biodiversity surrogate elaborated using abiotic variables to predict the presence of threatened species. Soil carbon could thus help to prioritize sites in conservation planning. Further, the results suggest that soil carbon on its own can be a much better surrogate than other abiotic variables when prioritization of sites for conservation are calibrated on increasingly small sets of observation plots. We call for the inclusion of soil carbon data in the elaboration of surrogates used to optimize conservation investments in the developing world.