High-frequency ultrasound standing waves (megasonics) have been demonstrated to enhance oil separation in the palm oil process at an industrial level. This work investigated the application of megasonics in the olive oil process on laboratory and pilot scale levels. Sound pressure level and cavitational yield distribution were characterised with hydrophones and luminol to determine associated physical and sonochemical effects inside the reactor. The effect of water addition (0%, 15%, and 30%), megasonic power levels (0%, 50%, and 100%), and malaxation time (10min, 30min, and 50min) was evaluated using response surface methodology (RSM) in a 700g batch extraction process. The RSM showed that the effect of the megasonic treatment (585kHz) in the presence of a reflector is more prominent at longer malaxation time (50min) and at higher water addition (30%) levels post-malaxation. Longer megasonic treatment of the malaxed paste (up to 15min; 220kJ/kg) increased oil extractability by up to 3.2%. When treating the malaxed paste with the same specific energy, higher oil extractability was obtained with longer treatments and low megasonic power levels in comparison to higher power levels and shorter times. Megasonic treatment of the paste before malaxation (585kHz, 10min, 146kJ/kg) and no water addition provided an increase in oil extractability of up to 3.8% with respect to the non-sonicated control. A double sonication intervention, before and after malaxation, using low (40kHz) and high (585kHz) frequency, respectively, provided up to 2.4% increase in oil extractability. A megasonic intervention post-malaxation (400 and 600kHz, 57-67min, 18-21kJ/kg) on a pilot scale using early-harvest olive fruits resulted in up to 1.7% extra oil extractability. Oil extracted under a high sonication frequency (free radical production regime) did not impact on olive oil quality parameters at reactor characterisation levels. Megasonic standing wave forces can enhance olive oil separation at various stages of the olive oil extraction process.