This prospective study explored whether extending the Theory of Planned Behaviour (TPB) with additional variables (descriptive norm, moral norm, anticipated affective reaction, self-identity and past behaviour) would account for additional variance in physical activity intentions and behaviour. Four interactions with past behaviour were also investigated. UK college employees participated in this study (n = 200, Time 1; n = 146, Time 2). Moral norm, self-identity and past behaviour each explained additional variance in intentions, over and above the TPB variables. Past behaviour moderated the impact of descriptive norm on intentions. Intentions, self-identity and past behaviour were significant predictors of behaviour, as measured by a self-report physical activity questionnaire. To increase physical activity, interventions might target moral norm and self-identity for physical activity. Focusing on positive descriptive norms might benefit sedentary individuals. The significant role of past behaviour is less useful in directing interventions.