ABSTRACT: Underwater noise has increased globally, yet our understanding of how baleen whales may be affected is limited. A controlled exposure experiment was used to determine whether an acoustic deterrent device (ADD) would modify the behaviour of minke whales Balaenoptera acutorostrata in Iceland. The Lofitech Seal Scarer in this study emitted a 14.6 kHz tone with a source level of 198 dB re 1 μPa re 1 m (rms). Video-range tracking was used to monitor behaviour during pre-treatment, control, treatment and post-treatment phases, with the ADD deployed at 338-1562 m during treatment (n = 10). Dive duration, swimming speed, reoxygenation rate and measures of path predictability were used to quantify responses. In all deployments, the ADD caused focal animals to move away immediately, typically increasing horizontal speed and directness away from the source and extending dives. Whales exhibited sustained movement away after deactivation. Deployment distance had the largest effect in a multiple regression model; closer deployments caused whales to move away faster. However, increases in dive duration and horizontal speed were apparently caused by the signal alone, rather than its magnitude. A noise propagation model suggested whales received maximum sound pressures of 150 dB re 1 μPa (rms) for the closest deployment (338 m). When modelling cumulative exposure over 24 h, deploying the ADD 25 m or more from individuals moving away at the average speed of 3.4 km h^-1 did not risk injury; temporary threshold shifts may be expected if deployed any closer. This study provides evidence that minke whales clearly react to signals at the likely upper limit of their hearing sensitivity.