The fundamental assumption in the management of predators is that reducing predator numbers will reduce their predation impact on livestock. Research on dingoes (Canis lupus dingo) has shown this assumption to be incorrect in beef production areas in northern and western Queensland. Aerial and ground baiting with 1080 (fluoroacetate) is the principal dingo-control method used in extensive pastoral areas of Australia. This paper compares four approaches to measure the effectiveness of these control programs. Dingo abundance was reduced in 11 of 13 baiting campaigns monitored with almost two-thirds of these producing >50% reduction. However, concurrent decline in dingo abundance occurred in non-baited areas due to seasonal changes in dingo populations. When this was taken into account, less than half of the control programs produced >50% reduction. The time taken for dingoes to recolonize baited areas is also an important measure of effectiveness. In two-thirds of the control programs, conducted in the first nine months of the year, dingoes recolonized prior to the period of peak calving (November/December) when the biggest threat to calves existed. The timing and the scale of control programs affect the rate of re-colonization. Calf loss was subsequently higher and occurred more frequently in baited areas compared to non-baited areas. Seasonal conditions, the status of prey populations, and the impact of control programs on social organization and prey selection, are key factors affecting calf predation. Control programs should be assessed by measuring impact rather than changes in predator numbers. The assumption that a direct relationship exists between predator numbers and impact is not valid for dingoes in beef production areas in northern Australia.