More than 18 000 archaeological sites have bxeen identified, and a dozen rock shelter sites investigated in the Zeekoe Valley, Upper Karoo, South Africa, but no work has been performed to date regarding edible plant remains. A baseline for understanding potential contribution of plants to the Zeekoe Valley diet is crucial for developing models of land use and mobility patterns for Late Stone Age (LSA) inhabitants, particularly since ethnographically known foragers in Botswana, some 800 km to the north, are as much as 80% dependent on plant foods for their survival. Since native foraging groups are long extinct in this semi-arid region with abundant natural springs, a rigorous investigation of botanical and ethnohistoric literature forms the groundwork for field investigations where direct ethnographic observation is no longer an option. First, edible species in the study area are identified in the literature. They are then sought on the ground with the aid of local informants whose families have resided in the valley for generations. It appears as though some knowledge of local plants gained from native foragers 150 years ago or more has remained in Afrikaans families passed down from landowning parent to child. On the ground, collection and middle range experimentation was followed by basic nutritional analysis. These measures were then used to compare foraging efficiency to measures from extant foraging groups. While dozens of edible species were identified, five plant species formed the focus of this study: Slymstock Uintjie (Albuca canadensis); Boesman Uintjie (Cyperus usitatus, Cyperus fulgens); Jakkalsbosberry (Diospyros austro-africana); Rooi wortel (Pelargorium sidoides); and Osbossie (Talinum caffrum). Results suggest native foraging groups, now extinct in the vicinity, may have been 60% or more dependent on plant food resources, despite abundance of game in the area and in the archaeological record.