Small cetaceans are susceptible to incidental mortality in the various forms of gillnet fisheries throughout their range. Research conducted since 1994 has shown that acoustic alarms (pingers) emitting high-frequency pulsed sounds effectively reduce the number of harbor porpoise Phocoena phocoena casualties in sink gillnets. However, the mechanisms behind the effects of pingers were still not understood. Until now, advantages and risks associated with their widespread use could not be evaluated. Here we present the results of 2 field experiments: (1) theodolite-tracking of harbor porpoises exposed to a single PICE-pinger in Clayoquot Sound, Vancouver Island, Canada and (2) herring Clupea harengus capture rates in surface gillnets equipped with and without acoustic alarms (Dukane Netmark 1000, Lien, PICE) in the Baltic Sea herring fishery at Rügen Island, Germany. Our results show that harbor porpoises do not seem to react to an experimental net in their foraging area (n = 172 groups, median group size = 2 porpoises). Porpoise distance from the mid-point of the net was distributed around a median of only 150 m (range 4 to 987 m). A net equipped with an acoustic alarm, however, was avoided (n = 44 groups) within audible range (distance distribution median = 530 m, range 130 to 1140 m). The porpoises were thus effectively excluded from the ensonified area. Herring, one of the main prey species of harbor porpoises, were not affected by the acoustic alarms tested (n = 25407 fish captured). The advantages and risks of using acoustic alarms to mitigate by-catch are discussed.