Dogs are the most significant threat to adult kiwi (Apteryx spp.), and kiwi aversion training (KAT) has been developed to reduce the likelihood of dogs harming and killing kiwi. In KAT, dogs are presented with a range of kiwi-related stimuli, which attempt to emulate live kiwi. The kiwi stimuli used can vary between trainers and may affect the overall efficacy of KAT. Thus, there is a need to assess the innate ‘interest’ of dogs in relation to kiwi stimuli to evaluate the potential relevance of different kiwi stimuli for KAT. In this study, I assessed three groups of dogs with differing experience with kiwi: kiwi conservation dogs, KAT dogs, and kiwi naïve pet dogs. I presented dogs with four kiwi-related stimuli: a kiwi carcass, a taxidermy kiwi, kiwi feathers, kiwi scats, and a blank stimulus using a single-choice preference assessment. Within each trial, dogs were presented with a single kiwi or blank stimulus and were allowed two minutes with the stimulus. The amount of time the dogs spent within the experimental area as well as the behaviour of dogs were videoed throughout the experiment. I used the amount of time the dogs spent investigating a stimulus, as well as their latency to approach the stimulus, as measures of ‘interest’ from the dogs to each stimulus. I found that, on average, dogs in all three groups spent more time investigating the kiwi carcass and taxidermy kiwi stimuli. However, the KAT dogs showed the lowest mean duration of investigating each stimulus, while the kiwi conservation dog group had the highest duration. For dogs in the naïve and kiwi conservation dog groups, there was a significant difference in the amount of time they spent investigating each stimulus, with the dogs spending significantly longer investigating the kiwi carcass and taxidermy kiwi in comparison to the scats, feathers and blank stimulus. There was no significant difference in the time the dogs spent investigating the different stimuli in the KAT dog group. Despite the decreased duration of investigation behaviour, dogs within the KAT dog group did not show fear-related behaviour in response to the kiwi stimuli presented. When the duration of time that the dogs spent investigating the different kiwi stimuli was compared across all three groups, no significant difference was found. This indicates that dogs across groups spent similar amounts of time investigating each stimulus. The lack of obvious learned avoidance behaviours towards kiwi stimuli by KAT dogs in this study suggest that dogs did not show a generalised avoidance toward kiwi stimuli that were not used in their initial KAT. Dogs may still show avoidance toward the stimuli used in their initial training but learned avoidance may not generalise to other kiwi stimuli even if these stimuli are the same type. This poses an issue as each trainer uses different stimuli during KAT, and if dogs are not learning to avoid all kiwi stimuli, dogs may not adequately learn to avoid live kiwi. The kiwi carcass and taxidermy kiwi were of the greatest ‘interest’ to dogs across groups and so these may be useful stimuli for use in KAT, more so that kiwi scats and feathers. From this study, it seems that visual cues provoke an initial ‘interest’ in the dogs to investigate a stimulus. I suggested to use several conspicuous items such as taxidermy kiwi models during KAT to ensure generalisation across different stimuli and increase the efficacy of KAT.