Harbour porpoise bycatch may be reduced by deterring porpoises from nets acoustically. In this study, two harbour porpoises were subjected to three acoustic alarms. The effect of each alarm was judged by comparing the animals' position and respiration rate during a test period with that during a baseline period. The XP-10 alarm produced 0.3 s tonal signals randomly selected from a set of 16 with fundamental frequencies between 9 and 15 kHz, with a constant pulse interval of 4.8 s (duty cycle 6Ž The 2MP alarm produced 0.3 s tonal signals randomly selected from a set of 16 with similar fundamental frequencies but with random pulse intervals of between 2 and 5 s (duty cycle 8Ž The frequency spectra and source levels of the 2MP and XP-10 alarms varied depending on the signal selected. The HS20-80 alarm produced a constant, but asymmetrical frequency modulated sinewave between 20 and 80 kHz with total pulse duration of 0.3 s, with random pulse intervals of between 2 and 5 s (duty cycle 4.6Ž The porpoises reacted to all three alarms by swimming away from them and by increasing their respiration rate. The XP-10, which on average had the highest source level, had the strongest effect.