Objectives: The purpose of this study was to investigate the effects of pet dog-assisted activities on self-esteem, depression and cognitive function of elderly people Method: The experimental group(pet dog-assisted group) consisted of 22 people over 65 years old out of about 90 elderly people registered with a home for the aged and the control group(non-pet dog assisted group) consisted of 22 people over 65 years old out of about 90 elderly people registered with another home for the aged in Daegu. Both groups had similar general characteristics, state of health, conduct on health and self-esteem, depression and cognitive abilities. Nineteen people who participated in the program at least 10 times out of 12, were chosen for the final analysis from the experimental group. For the control group, 20 people who responded to all three tests, which had been conducted before, 1 week after, and four week after the experiment, were analyzed. Result: Mean(${pm}SD$) differences between before and 1 week after the experiment for self-esteem, depression, and cognitive function were 5.84(${pm}3.06$), -3.26(${pm}1.76$), 1.47(${pm}1.26$), respectively. Those between before the experiment and 4 weeks after the experiment was 5.68(${pm}3.64$), -3.94(${pm}2.32$), 1.63(${pm}0.96$), respectively. Those for control group between before and after the experiment were -0.40(${pm}1.27$), 0.00(${pm}0.92$), 0.15(${pm}0.93$), respectively. Those for control group between before and 4 weeks after the experiment were -0.45(${pm}2.24$), 0.25(${pm}1.68$), 0.15(${pm}0.93$), respectively. All the differences between experiment and control group in mean differences between before and 1 week after experiment, between before and 4 weeks after experiment were statistically significant(p<0.05). Conclusion: This study suggests that the pet-assisted activities can be useful solution for elderly people who have psychological and emotional problems caused by retrogression of physical, mental and social ability.