Inactivation of Escherichia coli in liquids was carried out using power ultrasound. Parameters examined included amplitude levels (0.4 µm, 7.5 µm, 37.5 µm), treatment time, cell condition (non-adapted cells, acid adapted cells), liquid media (TSB, model orange juice and model apple juice) and E. coli strain (ATCC 25922, NCTC 12900). The efficacy of ultrasound treatment was found to be a function of amplitude level, treatment time and media (p < 0.05). The kinetics of inactivation followed zero order kinetics (R > 0.95), with the highest inactivation achieved using an amplitude of 37.5 µm. The D-values of E. coli 25922 at all amplitudes in model orange juice were not significantly different than in TSB media. However, at 0.4 µm and 37.5 µm amplitude D-values of E. coli 12900 were significantly different in model orange juice compared to TSB media. When efficacy of ultrasound was assessed in model apple juice and phosphate buffered saline treatment times were significantly reduced by comparison with TSB. Inactivation of E. coli was found to be influenced by strain, prior acid adaptation and suspension liquid, but the effect was negated at the higher amplitude levels.