Subsurface erosion in the loessial soils of Banks Peninsula has been studied by numerous researchers and the mechanism and form of the erosion process is now w ell understood. This thesis study is a practical evaluation of potential investigation methods at sites of subsurface erosion, with the additional evaluation of suitable remedial options for erosion stabilisation. Preferred laboratory procedures include the pinhole test, the crumb test and the uniaxial expansion test. While the pinhole test is the most useful single measure of erodibility, the crumb test is the more effective test of dispersion. When the results are combined, these three tests appear to give a reliable guide to soil material erosion potential. Soil mass layering and defects also provide important controls on subsurface erosion, and auger and testpit logging allows soil mass characterisation. Geomorphic mapping provides the first approximation to the location, extent and form of subsurface erosion. Photographic methods, as aids to this mapping, appear to have only limited application. Geophysical methods are not able to define cavity location or extent but are still useful as general site investigation tools. Dyed water and smoke can be used to trace cavity connections and establish tunnel outlets. The interception of surface before it enters the area of principal the cheapest remedial technique. or subsurface water concern is generally If the subsidence or collapse of large cavities creates problems then a number of options are available. Foundations at risk can be supported on piles extending through the erosion cavity to bedrock, but in accessible areas excavation to below cavity level with drainage and chemically stabilised backfilling has proven effective. The slurry filling of subsurface cavities with lime/cement/sandy gravel slurries appears a promising new method and the application of this technique at one site has provided considerable cost savings.