Pre-school children are at the highest risk of being bitten by dogs, and previous research suggests that this is due to limitations in their emotion recognition ability. The use of stimuli in emotion recognition literature has been criticised for being limited by using photographs as videos acquire better performance. 25 pre-schoolers from Edinburgh aged 3 – 5 years old (mean age=4.1 years) were shown photos and videos of dogs and humans displaying the four basic emotions (anger, happiness, sadness and fear) and were asked how they thought the dog or human was feeling. They were then asked what body parts they were looking at to be able to tell this, and why the child thought they could be feeling this way. Their attitude towards dogs and ability to understand that dogs have emotions was also compared to their dog emotion recognition performance. T-test comparisons found that pre-schoolers performed significantly better in the human (M=87%) than dog condition (M=43%), but stimulus type had no effect on performance. Children mainly attended to the face when looking at humans but looked at the whole body when recognising emotions in dogs. Attitude and ability to understand that dogs have emotions had no effect on ability to recognise dog emotions. Experience with dogs did not seem to have an effect but sampling limitations did not allow for inferential testing. The addition of sound to the stimuli in the future could aid video emotion recognition. If this did improve performance it could be helpful in dog bite prevention programmes and may in turn reduce the number of dog bite incidents. Similarly, if eye-tracking studies were carried out on similar stimuli, prevention programmes could assist children in learning where they need to attend to in order to recognise a dog’s emotion by focusing upon their weaknesses.