Summary1. Toxicant-resistance is a potential, or very real, problem with many pest-control programmes world-wide. However, apart from rodents, pesticide-resistance has not been well documented in vertebrates. We assessed the potential impact of developing resistance to 1080 in rabbit populations with differing levels of historical exposure to 1080-baiting programmes in south-western Australia. 2. The sensitivity to 1080 of three out of the four populations of rabbits Oryctolagus cuniculus examined had decreased significantly since Australian rabbits were last tested over 25 years ago. The lethal dose 50 (LD 50 ) values for these populations, as determined from formal toxicity trials, ranged from 0·744 to 1·019 mg pure 1080 kg − 1 , and were significantly greater ( P < 0·05) than the previously reported values for Australian rabbits (LD 50 range 0·34-0·46 mg pure 1080 kg). The LD 50 value for the fourth population (0·584 mg pure 1080 kg − 1 ), which has had the least exposure to 1080, did not differ from that reported previously ( P > 0·05). 3. The lethal dose 99 (LD 99 ) values for the four rabbit populations tested ranged from 1·181 to 1·666 mg pure 1080 kg − 1 , and suggested that, theoretically, all rabbits should be killed during routine baiting campaigns provided that there is no loss of active ingredient from the bait. In reality, the efficacy of 1080 poison bait laid in trails for controlling free-ranging rabbits was reduced in those populations where rabbits had decreased sensitivity to 1080. Mean reductions in rabbit numbers 7-9 days after trail baiting of resistant and sensitive populations ranged from 51·2% to 65·2%, and from 76·4% to 76·5%, respectively. 4. These findings suggest that genetic resistance to 1080 is developing in at least some populations of Australian rabbits. This has world-wide implications for agricultural protection and wildlife conservation programmes that rely on a 1080-baiting strategy for reducing the impact of vertebrate pests.