Underwater acoustic pingers, emitting a 10 kHz sound (with harmonics up to 110 kHz), were tested in New Zealand to evaluate their potential effectiveness in preventing entanglement and death of Hector's dolphins (Cephalorhynchus hectori) in gillnets. A 1996 study site was established in Akaroa Harbour, an area known for high concentrations of Hector's dolphins. A remote controlled device was installed which would raise and lower either active or passive acoustic pingers via a radio link from shore. Observations of dolphin movement and distribution were made from a land-based station using a theodolite and logged directly into a computer. Observers did not know if the pinger in the water was active or passive. This blind experiment was designed to measure the spatial difference in dolphin distributions between active or passive pinger use. Two data subsets were used in the analysis, representing the distance between sighted dolphins and an active pinger and the distance between sighted dolphins and a passive pinger. The distribution of the distance data was significantly non-normal (p < 0.001), so the non-parametric Mann-Whitney rank sum test was used to compare dolphin distributions from the two subsets. In this analysis, all sightings data were included; the median distance value for the passive pinger trials was 299m (n = 492) and the median distance value for the active or ensonified trials was 372m (n = 552). Results indicate that Hector's dolphin distributions were affected by the 10 kHz pingers and that dolphins avoided the immediate area where the pingers were active, but did not avoid the larger area of Akaroa Harbour. All dolphin sightings made during active pinger trials were distributed significantly farther from the sound source (p < 0.001) than were sightings during passive trials.