Soil properties are indicators for ecological processes and thus contribute to determining "functional and self-sustaining ecosystems" in a rehabilitation context. In a recovering ecosystem these indicators are expected to follow a trend towards a benchmark. Whether such a trend can be observed in rehabilitation projects in an arid environment was the question of this study. Soil properties of restored areas with six different treatments and corresponding reference sites were analysed at Trekkopje Mine in the central Namib Desert over six years. Soil properties which were reasonably stable over the monitoring period in reference sites, and not even affected by rainfall patterns, were pH, organic carbon, calcium, potassium, magnesium and clay content. The chemical indicators were likely linked to the treatments, although clear patterns had not yet developed. Organic carbon content was, however not linked to treatment or standing biomass. The best re-vegetated sites showed the lowest organic carbon, and thus no link between standing biomass and soil organic carbon. This may indicate that factors other than standing biomass control soil organic carbon and therefore call into question its use as an indicator of soil fertility in arid, recovering ecosystems. Control, scarified and topsoil-treated sites showed a clear trend in declining calcium, possibly as a result of the exposed, initially highly calcareous subsoil and subsequent leaching. Therefore, only one short-term soil indicator was supported by this study and more time and possibly a larger sample size are needed to show trends in other soil properties. Long-term data collection which consistently applies the same monitoring protocol is therefore essential in an arid environment and longer time intervals between monitoring events (e.g. 2-3 years) can be considered, if costs need to be reduced. Keywords: completion criteria, mining, Namib Desert, rehabilitation, soil fertility, substrate treatments.