It is known that pet dogs can act as catalysts for human social interactions, and it has been suggested that this may enhance feelings of well-being. Two studies were carried out to establish the robustness of this effect. In Study 1, a highly trained dog was used to ensure that the dog itself did not solicit attention from passers-by, and data were collected across a range of normal daily activities in which a dog could be included, not confined to conventional dog walking areas as in previous studies. Being accompanied by a dog increased the frequency of social interactions, especially interactions with strangers. In Study 2, also using a trained dog, a different (male) participant observer was dressed either smartly or scruffily. Although there were significantly more interactions when he was smartly dressed, the greatest effect was between the Dog present and No Dog conditions irrespective of the handler's dress. It is concluded that the social catalysis effect is very robust, which opens the way for investigating possible consequences of the effect for well-being and health.