1. Wild predators sometimes kill livestock. In Australia the red fox kills lambs, but there are limited experimental data to demonstrate the effects of controlling foxes on the predation level.
2. Using a balanced experimental design over 2 years, we investigated the effects of three levels of fox control, using poisoning with sodium monofluoroacetate (compound 1080), on lamb production, lamb predation and fox abundance in south-eastern Australia.
3. There was no effect of fox control on lamb production. Fox predation was the probable cause of lamb death for a minimum of 0·8% and a maximum of 5·3% of 1321 lamb carcasses. Fox control significantly (P < 0·05) reduced the minimum percentage of lamb carcasses classified as killed by foxes from 1·50% (no fox control) to 0·90% (fox control once per year) or 0·25% (fox control three times per year). Fox control also significantly (P < 0·005) reduced the maximum percentage of lamb carcasses classified as killed by foxes from 10·25% (no fox control) to 6·50% (fox control once per year) or 3·75% (fox control three times per year). Poisoning did not affect fox abundance in spring.
4. We estimated the number of treatment replicates needed to detect an effect of predator control on the number of lambs successfully reared to lamb marking (10 weeks old). The estimated numbers were high if predation effects were small. It is recommended that the effects of fox control should be evaluated carefully to determine if the present results are widespread, as some fox control may be wasted.
5. The general effects of predator control on livestock predation requires more experimental investigation and economic evaluation. However, field experiments of sufficient power to detect effects will present considerable challenge.