Depletion of shallow ore reserves is forcing the South African gold mining industry to exploit reefs at ever-increasing depths. By 2010 around 30 per cent of South African gold production probably will come from depths greater than 3 km (Johnson and Schweitzer, 1996). At such depths, virgin stresses will rise significantly, producing complex rock engineering problems (Malan and Basson, 1998). Concern over structural instability at depth has led planners to seek a technology capable of detection of geological structures with a displacement of 2 m or greater (or as appropriate to the mining technique) 200 m ahead of mining at an acceptable cost'. Borehole radar probably is such a technology. Cover holes are drilled routinely ahead of development. The evaluation of its candidacy for a tactical role in ultra-deep gold mining depends-among other things-upon establishing the electromagnetic characteristics of the environment hosting significant gold reefs. The Ventersdorp Contact Reef (VCR) is one such reef. Recent borehole radar field trials, presented here, suggest that the VCR's overburden is a remarkably translucent, clutter-free dielectric with a Q of ~35 and that the VCR itself is the dominant radar reflector.