In 2001, Mozambique declared an area known as "Coutada 16" (hunting zone) the Limpopo National Park (LNP), which forms part of a trans-frontier park with South Africa and Zimbabwe. The park provides ecosystem services and supports the livelihoods of thousands people living in the many communities within its boundaries, which were planned for relocation outside the park. These moves were expected to result in major land use changes, both in terms of vegetation and wildlife, affecting soil quality in and around the LNP, including in resettlement areas. Therefore, this study aimed at estimating soil organic carbon (SOC) stocks in the Limpopo National park as an indicator of livelihoods and ecosystem function. The estimation of SOC stocks was first attempted from legacy data then by both measure-andmultiply and digital soil mapping (DSM) based on a new sampling plan. Along this study issues of legacy data renewal quality, mapping data-poor and poorly-accessible areas, time-consuming and costly traditional method for SOC laboratory determination as well as uncertainty and reliability of SOC stocks estimates from different methods were also investigated. Overall this research provided (1) a guiding framework and quantitative measures for evaluation of renewed legacy survey and as such enabled an informed qualitative and quantitative SOC stocks estimation, (2) a cost-effective methodology for mapping SOC in data-poor, poorly-accessible areas following a DSM approach, (3) a rapid, cost-effect, non-destructive and pollutant-free Near-Infrared a calibration model for the determination of SOC in LNP and, (4) SOC stocks estimates, its uncertainty and reliability. Despite the high uncertainty of the estimates, which limit its use in baseline studies, achieved SOC stocks estimates are, in general, consistent with of SOC stocks estimates in the literature for similar soils in comparable environmental conditions in southern Africa.