The importance of regular and life long physical activity, both from a physical and medical perspective and from a psychological well-being perspective, is well documented. Also well documented is the reduction, below sufficient levels, in physical activity participation, particularly for populations such as young and mid-life adult women. Physical activity promotion is, thus, of great importance in modern society. In this thesis, my primary aim is to develop ways to enhance the adoption and maintenance of physical activity in young and midlife women. To do so, I have utilised the theoretical conceptions from a dominant social cognitive model, the theory of planned behavior (Ajzen, 1985). Subjective norm has not been a strong predictor of intention or behaviour in research on the theory of planned behavior in physical activity contexts (Symons Downs & Hausenblas, 2005), the reasons for which have been the subject of much debate in the literature. To date, there is disagreement about the conceptualization, measurement, and importance of the variable. In this thesis, I examined the potential of the social identity approach as a basis for making subjective norm more meaningful. In this approach, rather than assessing the influence of a random collection of important others as in the original definition (Ajzen, 1985), subjective norm is conceptualised as the perceived pressure from specific reference groups relevant to self, a conceptualization more in keeping with the social identity approach (Terry & Hogg 1996). I expected that the social identity approach to subjective norms would enhance its ability to predict intentions and be the basis of an effective physical activity intervention. Specifically, I expected that physical activity norms, when presented to women from members of a shared social identity would influence both intentions and behaviour.ii In the first study, I surveyed 214 women between the ages of 25 and 45 to obtain a list of the potential reference groups that would be relevant for young and midlife women. I asked women to list their self-characteristics and then to rate them on Simon's (1997) social self-categorisation index. I sorted the self-characteristics into types, and listed the frequency and social identity potential for each type. From the most frequently listed self-characteristic types, I found that the personality-based selfcharacteristics, such as strong independent woman, and spiritual caring woman, had more social identity potential, than the role-based self-characteristics, such as wife or professional. I used these personality-based self-characteristics as the basis for my intervention study.In the second study, I conducted a 3-month physical activity study based on the social identity approach to subjective norm. The intervention was for sedentary women (25 to 45 years), who engaged in less than 150 minutes of exercise per week. In two social identity conditions (SI), I conducted a social identity-based subjective norm manipulation. The first SI condition was for women (...