Interactive technology for dogs is on the rise with there now being a whole TV channel supplying videos for dogs as well as several examples of interactive technology designed for the purpose of ‘entertaining’ dogs and other pet mammals. The Animal-Computer Interaction (ACI) research field has moved into the study of such technologies having earlier focussed mainly on technologies for working animals where there are two main stakeholders, the dogs and the humans, and where there is an expectation often, that the dog is doing a certain task. In studies of what might be referred to as entertainment interaction, there is a need to understand what, if anything, attracts a dog's attention whilst being mindful that where the only real stakeholder is the dog, there may be some methodological considerations in regards to the level of control and the level of autonomy given to the dog in such studies. This paper presents a study of dogs’ attention between three screens to explore the movement of attention between screens, and between videos in a relatively uncontrolled research environment. The study demonstrates that the method used, which was to be as ‘dog-centred’ as possible, yielded useful data. For the design community it is shown that the dogs were seen to attend mainly to a favoured screen (left and centre in this case) and three of the videos appeared to be preferred over the others.