An in situ field technique has been developed and implemented to measure non-aqueous phase liquid (NAPL) entry pressure and subsequently estimate mechanical aperture variability in an undisturbed, water-saturated rock fracture. The field experiment also provided the opportunity to measure wetting phase relative permeability at residual non-wetting phase saturation, k(w,r)(S-nw,(r))The radial aperture determination by the injection of oil (RADIO) method employs a constant-rate injection of a non-toxic NAPL into a fracture isolated by a double packer array. Injection pressures are recorded to provide a characteristic signature that is interpreted to reveal aperture enlargements and constrictions several metres from the borehole. Fracture aperture values are calculated by reverse modelling of the pressure signatures using the FRACAS fully compositional two-phase flow model. Pre-NAPL injection hydraulic testing established a baseline dataset of hydraulic apertures ranging from 26 pm to 259 pm with a mean value of 86 pm. Post-NAPL injection hydraulic testing gave k(w,r()S(nw,r)) values ranging from 0 center dot 04 to 0 center dot 39 with a mean value of 0 center dot 15. Numerical modelling calculated fracture mechanical apertures vary from 36 mu m to > 1000 mu m in test 1, and from 46 mu m to > 1000 mu m in test 2. Single-value approximations of volume-averaged fracture apertures are two to three times larger than the respective hydraulic apertures, suggesting that volume balance calculations using the hydraulic aperture may overpredict NAPL penetration distances.