Grazing livestock freely select landscape resources, unless they are herded or constrained by fences. Automated animal control (AAC) systems offer an alternative to physical fences by using animal-positioning technology and aversive stimuli to deter animals from staying in sensitive environments and so limit their impact. This paper reports on a replicated field experiment completed to test whether occasional stimuli (audio cue followed by a mild electric stimulus), delivered by discontinuously activated AAC collars, could suffice to modify the grazing behavior of groups of cattle. Four groups of eight steers were confined in 8-ha rectangular paddocks that had an ad libitum supplement feeder located in one end to attract cattle. The steers' positional information was recorded continuously for 3 d using a GPS receiver encased in a collar fitted around their neck. These data were used to characterize their use of the paddocks without intervention. Subsequently a restriction zone was activated on the collars. This zone contained the supplement feeders and represented approximately 10% of the paddock area. Cattle movement was again monitored during a second 3-d period, in which the steers were subjected to discontinuous aversive stimuli (5 min of stimulation followed by a random 0–30 min interval without stimulation) if they were located inside or moved into the restriction zone. Cattle visits to the restriction zone were shorter and the return interval longer when steers were subjected to discontinuous stimulation. Overall, there was a 97% reduction in the use of the restriction zone between the first and second deployments. These results suggest that grazing impact can be drastically reduced by making a zone less desirable through discontinuous aversive stimulation. Such a discontinuous (25% of the time on) AAC system can reduce power consumption in collars and so help overcome energy supply limitations that hinder commercial AAC applications.