The hafting of tools using adhesive is one of the innovative features that characterizes the southern African Middle Stone Age. This technology has mainstream implications but remains insufficiently documented, largely due to unequal organic preservation and non-adapted analytical procedures. A notable exception is provided by the recent results from the site of Sibudu (Lombard, 2006;Wadley et al., 2009). The excavation at Diepkloof Rock Shelter has revealed several lithic artifacts with a black residue distributed over their surface. Their stratigraphic distribution reveals a strict association with the Howiesons Poort (HP) and suggests a close relationship between the appearance of hafting adhesive and the appearance of blades and geometric backed tools. Macroscopic and microscopic observations attest to a hafting that was exclusively lateralized and preliminary use-wear analysis (Igreja and Porraz, 2013) supports the hypothesis that hafted tools were mostly integrated within daily (domestic) activities. In this study, we focused specifically on a chemical study of a thick black residue found on a quartz flake attributable to the Late phase of the HP. We determine, for the first time in a MSA context, the nature of the compound adhesive and reconstruct a picture of the multilevel operations and interactions that comprise the process of hafting. The molecular analysis attests to the exploitation of Podocarpus elongatus (Yellowwood), collected in the form of a resin that was naturally dried or heated at a low temperature and possibly mixed with fragmented bone and quartz grains. Compared to Sibudu where ochre additive is documented, the hafting technology at Diepkloof introduces another level of variability within the HP tradition and suggests the existence of regional expressions and adaptations.