This dissertation explored the possibilities and challenges of using a social game in conjunction with selected members of personal communities to promote physical activity. Digital entertainment media enjoyment and the role of close social ties as partners for health promotion were investigated through the evaluation of an intervention called Wellness Partners. Mixed methods of an experiment of pretest-posttest control group design and qualitative analysis of server log files as well as in-depth interviews were used. The study found that the average user managed to log in almost every other day, some with more enthusiasm than others. Quantitative analysis suggested that their enjoyment and participation did not differ significantly between the experimental and control conditions both at the individual and group level. On average, 3.47 partners were nominated and 1.60 were able to enroll in the study, with over two fifths from immediate family and the rest from close friends and colleagues of shared interests. Members of larger study groups participated significantly more although they didn’t necessarily enjoy it better. No statistically significant health effects were found in the quantitative analysis regarding exercise self-efficacy, perceived social support, exercise habits, and perceived wellness. However, qualitative analysis indicated that participants used the website differently and three major web activities were tracking, socializing, and gaming. Their comments about the most and least enjoyable experiences supported a three-level enjoyment hierarchy model. The results suggested that overall participants encountered a lot of technical difficulty, yet still they appreciated the social and gaming features and demanded more. The intervention made some qualitative differences in participants’ perceptions and behavior related to physical activity and holds great potential if the play activities are tied to specific health goals and content.