European Red Fox (Vulpes vulpes) baiting with 1080 poison (sodium fluoroacetate) is undertaken in many Australian sites to reduce fox abundance and to protect vulnerable native species from predation. The longest continuous use of fox baiting for fauna conservation commenced in south-west Western Australia in the 1980s and includes baiting Dryandra Woodland and Tutanning Nature Reserve. The trap success of the Woylie (Bettongia penicillata) in these two reserves initially increased more than 20-fold after the commencement of baiting and was maintained until 2000. Woylie captures then decreased rapidly, despite ongoing fox baiting, so the long-term efficacy of 1080 baiting was questioned. Here, fox density and probabilities of detection, re-detection and survival between replicated baited and unbaited sites were compared by modelling capture–recapture of individual foxes. These were identified from microsatellite DNA genotypes obtained non-invasively from hair, scat and saliva samples. The frequency and duration of fox residencies were also quantified. Remote cameras were used to determine the fate of baits but uptake by foxes was low, whereas nontarget species' bait uptake was high. Nevertheless, foxes inhabiting baited reserves had significantly higher mortality, shorter residency times, and 80% lower density than foxes inhabiting unbaited reserves. Baiting continues to significantly reduce fox abundance after more than 25 years of continuous use. This has positive implications for fox control programmes throughout Australia but reduced fox abundance may facilitate increased predation by feral Cats (Felis catus).