Since the return of the wolverine Gulo gulo to the Snøhetta area in Norway in 1979, wolverine predation on sheep Ovis aves has increased in parallel with increases in the number of sheep. Reducing the predation would also reduce the economic losses suffered by sheep farmers and could increase their willingness to accept the presence of wolverines in the area. Therefore, experiments with volatile repellents to reduce predation by wolverines on sheep were carried out. Experiments showed that five oils and three pure chemicals gave distinctive avoidance reactions by captive wolverines. The release rates of the different chemicals were tested in the laboratory and a dispenser allowing the use of the chemicals as long-lasting repellents was developed. In 1993, half of the ewes and their lambs in areas with high losses due to wolverine predation were randomly chosen and fitted with dispensers containing these chemicals. In 1994, the test was repeated and the study area was expanded to include a nearby area, in which farmers had also suffered high wolverine predation on their sheep in recent years. All sheep flocks were monitored during the free-ranging grazing period, and when carcasses were found, necropsies were performed to ascertain the cause of death. Nearly all dispensers were damaged to some degree during the grazing period, but because family groups (ewes and their lambs) were marked, this is believed to probably not have affected the experiment significantly. In spite of the technical failures, significantly fewer losses occurred in the groups with dispensers than in the groups without dispensers. A better dispenser must be developed and a large-scale test should be performed before it is possible to conclude definitively that volatile repellents may be used as an operational instrument to reduce economic losses to sheep farmers and to help conserve wolverines.