Sorghum is a known reservoir of drought adaptation mechanisms, but little detailed attention has been paid to root system attributes. In addition to improving drought adaptation, recent agronomic practices of varying row configuration (e.g. skip row systems) might be enhanced by improved understanding of root system architecture. We sought to explore extent of soil water capture by sorghum, its spatial variability in the soil, and its genetic regulation via factors affecting root system architecture. A range of field and rhizotron studies were conducted.  An initial field study indicated water extraction progressed out from the plant radially to a distance beyond 2m. Nodal root angle was hypothesised as an indicator of genetic variability in root system architecture and a phenotyping system using mini-rhizotrons was developed. Significant variation in nodal root angle was identified across a diverse range of sorghum inbred lines. Four inbred lines with contrasting root angle were grown in large rhizotrons to study their water extraction patterns. An association between nodal root angle and the spatial distribution of the root system and water extraction of mature plants was observed. QTL for nodal root angle were mapped by screening a population that was developed by crossing two inbred sorghum lines with contrasting root angle. The four nodal root angle QTL identified explained 58.2% of the phenotypic variance and were validated across a range of diverse inbred lines. Three of the four nodal root angle QTL showed homology to previously identified root angle QTL in rice and maize, whereas all four QTL co-located with previously identified QTL for stay-green in sorghum. A putative association between nodal root angle QTL and grain yield was identified through single marker analysis on a subset of the mapping population grown in field trials in hybrid combination with three different tester lines. The identification of effective QTL associated with root system architecture presents new opportunities for improving drought adaptation via molecular breeding.