Incidental bycatch in fishing nets is a global cause of incidental mortality of marine mammals. Two classes of approaches attempt to mitigate this impact: (1) approaches that change the behaviour of the fisher (e.g. closures and gear modifications), (2) approaches that attempt to change the behaviour of the bycatch species (e.g. acoustic alarms or pingers). Even though the effectiveness of pingers has been established for very few bycatch species, pingers are now mandatory in many fisheries throughout the world. Pingers are being trialled in commercial gill net fisheries in tropical Australia to reduce the bycatch of the dugong and three species of coastal dolphins, despite an absence of robust assessments of: (1) their effectiveness in reducing bycatch, (2) the likelihood of alienating bycatch species from critical habitats. We conducted replicate experiments to test the behavioural responses of dugongs to 4 and 10 kHz pingers in an array simulating a net. Each experiment comprised three sequential 10-min treatments in which two pingers were: (1) inactive, (2) active, (3) inactive. The rate of decline of the number of dugongs within the focal arena did not change significantly while pingers were activated. Dugongs passed between the pingers (where a net would be located) irrespective of whether the alarms were active or inactive, fed throughout the experiments and did not change their orientation to investigate pinger noise, or their likelihood of vocalizing. We conclude that: (1) pingers are unlikely to alienate dugongs from critical habitats or reduce dugong mortalities in fishing nets, (2) bycatch mitigation strategies such as pingers that rely on changing animal behaviour should only be used after rigorous testing on all likely bycatch species.