A randomised controlled trial was used to evaluate a theory-based health promotion intervention. The intervention, a healthy living booklet, was designed to promote healthy eating and physical activity amongst people aged over 65 years attending hospital out-patient clinics. The booklet employed persuasive arguments targeting the most proximal cognitive antecedents of behaviour specified by the theory of planned behaviour, as well as goal setting prompts. Participants (N = 252, average age=82) were randomly allocated to a control (patient satisfaction questionnaire) or intervention (healthy living booklet) group. Cognitions and behaviour were measured pre-intervention and at a two week follow up. The intervention group made significantly higher gains in perceived behavioural control, intention and behaviour for both target behaviours, suggesting that the intervention was successful. Sixty three of those invited to set goals to eat more healthily (e.g., "to eat five portions of fruit and vegetables a day") did so, and 67% of those who set such goals reported 100% success in acting on them. By contrast, only 34% of intervention participants set an activity goal (e.g., "a five minute walk everyday"), and only 51% reported 100% success in enacting these goals. Results suggest that the observed behavioural effects of the healthy eating booklet could be attributed to goal setting as well as changes in perceived behavioural control and intention.