This study aims at testing long-term learning effects of previous sheep tests on sheep chasing in hunting dog breeds (Norwegian Elkhounds, grey, English setters, and hare hunting dogs), in particular with use of electronic dog collars, in addition to uncovering potential secondary negative effects on dogs’ behaviour and mental stability. The dogs (N=114) were subjected to three tests for two subsequent years, the second year being reported here. Dogs were tested for reactions to different stimuli, including a sheep, in a path test. In a sheep confrontation test dogs were fenced in with a sheep group and given el. shocks when approaching 1-2 m from sheep. A questionnaire to the dog owners reported differences in dogs’ behaviour between the years. Dogs showed weaker or delayed behavioural responses in both tests in the second year. No dogs showed interest in or attacked a lone sheep in the path test in the second year, while almost two thirds of them did so the first year. In the sheep confrontation test, the dogs exhibited comparatively hesitant initial hunting motivation the second year, being more evident in dogs which received el. shocks the first year. No dogs chased or attacked sheep as their first response in this test, while half of them did so the first year. The proportion of dogs attacking sheep during the entire test was reduced to almost one fourth. The number of el. shocks administered reduced by the second year, and only one dog that received el. shocks the first year received el. shocks the second year. The second-year tests indicate that aversive conditioning with the use of electronic dog collar may be an efficient method for reducing the probability of a dog chasing or attacking grazing sheep. No adverse effects were observed with our test procedure.