The main aim of this study was to evaluate pet therapy in terms of increased interaction with two groups of elderly psychiatric patients. These patients were different in respect of their diagnosis, but their environments were similar in layout, decor and facilities. Each group of patients was observed by an independent observer, and further rated by the ward nurses on a scale which recorded change in their behaviour over the three conditions: baseline, dog intervention and photographic intervention. The total time of observing was four weeks where the baseline (i.e. “normal” amount of interaction) and the two intervening stimuli were introduced, and compared with one another. Time sampling was used throughout the study. The results indicate that the dog intervention was the most effective in increasing interaction when comparisons were made with other conditions.