Congestive heart failure (CHF) is one of the leading causes of hospitalisation, morbidity and mortality in the UK. The incidence and prevalence of CHF is expected to increase due to the aging population and improved survival in heart disease. Exercise has been recognised as a valuable treatment and has proved to be beneficial in CHF. The present study applied the Theory of Planned Behaviour (TPB) and the Commonsense Model of Self-regulation to fitness, daily activity and medication-adherence in elderly patients with CHF. The study was in parallel with a randomised controlled trial of a 3-month exercise programme. A TPB questionnaire was used at baseline and at 3 months. Illness representations were assessed by IPQ-R at baseline. Fitness (measured by 6 minute walk test) and daily activity (measured by an accelerometer) were assessed at baseline, 3 and 6 months. ACEI-adherence (assessed by measuring serum ACE level) was assessed at baseline. The main findings from 81 CHF patients (mean age = 81 years) showed that Subjective norm (SN) was the only predictor of Intention (IN1) at baseline; Attitude, Perceived behavioural control (PBC) were predictors of INT at 3 months; fitness was predicted by PBC at 3 months. The exercise intervention led to significant changes in Attitude and PBe. The finding also indicated that the participants were more likely to attribute their illness to Chance, BiolOgical factors and God. Identity and Illness coherence predicted fitness, and Consequences predicted daily activity. Participants who believed that their illness was chronic or serious were less likely to adhere to ACEI medication. Conclusion: both the TPB and IPQ-R were useful instruments to predict behaviours in elderly patients with CHF. The IPQ-R had a greater predictive power than the TPB in this population. Illness representations may play a role in influencing the formation of intention as background factors.