Vessel noise is a primary driver of behavioural disturbance in cetaceans, which are targeted during whale-watch activities. Despite the growing, global effort for implementing best-practice principles, to date, there are no regulations on whale-watch vessel noise levels. Here, we test the hypothesis that a whale-watch vessel with a low noise emission will not elicit short-term behavioural responses in toothed whales compared to a vessel with a louder engine. We measured behavioural responses (n = 36) of short-finned pilot whales (Globicephala macrorhynchus) to whale-watch vessel approaches (range 60 m, speed 1.5 kn). Treatment approaches with a quieter electric engine (136-140 dB) compared to the same vessel with a louder petrol engine (151-139 dB) (low-frequency-mid-frequency weighted source levels, re 1 mu Pa RMS @ 1 m) were examined. Focal whales were resting mother and calves in small group sizes. During petrol engine treatments, the mother's mean resting time decreased by 29% compared to the control (GLM, p = 0.009). The mean proportion of time nursing for the calf was significantly influenced by petrol engine vessel passes, with a 81% decrease compared to the control (GLM, p = 0.01). There were no significant effects on behaviour from the quieter electric engine. Thus, to minimise disturbance on the activity budget of pilot whales, whale-watch vessels would ideally have source levels as low as possible, below 150 dB re 1 mu Pa RMS @ 1 m and perceived above ambient noise.