Studies revealed that primates are able to show visual preferences for particular species in free choice tests. As it also was shown that dogs possess the ability to discriminate and recognize photographs on a screen, this study investigated if they prefer particular visual stimuli. For this purpose 30 domestic dogs were tested in an experimental two-choice task on a touch screen on which six different stimulus combinations were shown: own breed vs. unfamiliar breed, familiar breed vs. unfamiliar breed, breed with an ancestral type face vs. breed with wrinkled face, dog vs. human, dog vs. other mammal and control set: appetitive objects vs. non appetitive objects. They were rewarded irrespectively of their choice. Even though individual preferences occurred quite rarely throughout the six sets, group level preferences could be detected in three of them. Dogs preferably touched photographs of unfamiliar breeds which can be explained by a dishabituation effect or visual preference for novelty out of curiosity. On the other hand, an additional analysis revealed that some confounding parameters (body size, gaze direction and coat colour of the dog in the photograph) might have influenced subject’s choices. Interestingly, subjects preferred to choose dogs with averted gaze, which is known as a calming signal in real life. The tested group also touched pictures of dogs more frequently than pictures of humans. This suggests that social experience does not affect visual preference in dogs but the phylogenetic distance. Although this experiment provides evidence of some visual preferences in dogs, results should be discussed with caution as the orientation of these preferences remains uncertain.