Heat treatment was one of the first transformative technologies in the southern African Middle Stone Age (MSA), with many studies in the Cape coastal zone of South Africa identifying it as an essential step in the preparation of silcrete prior to its use in stone tool manufacture. To date, however, no studies have investigated whether heat treatment is necessary for all silcrete types, and how geographically widespread heat treatment was in the subcontinent. The aim of this study is to investigate experimentally whether heat treatment continued further north into the Kalahari Desert of Botswana and northernmost South Africa, the closest area with major silcrete outcrops to the Cape. For this we analyse the thermal transformations of silcrete from both regions, proposing a comprehensive model of the chemical, crystallographic and `water'-related processes taking place upon heat treatment. For the first time, we also explore the mobility of minor and trace elements during heat treatment and introduce a previously undescribed mechanism - steam leaching - causing depletion of a limited number of elements. The results of this comparative study reveal the Cape and Kalahari silcrete to respond fundamentally differently to heat treatment. While the former can be significantly improved by heat, the latter is deteriorated in terms of knapping quality. These findings have important implications for our understanding of the role of fire as a technical solution in MSA stone tool knapping, and for the extension of its use in southern Africa. Silcrete heat treatment - at least in the form we understand it today - may have been a strictly regional phenomenon, confined to a narrow zone along the west and south coast of the Cape. On the basis of our findings, silcrete heat treatment should not be added as a new trait on the list of behaviours that characterise the MSA of the southern African subcontinent.