INTRODUCTION: Weinman, Petrie, Sharpe, and Walker (2000) showed that the causal attributions of a sample of first-time myocardial infarction (MI) patients and their spouses from Auckland, New Zealand, were associated with changes in health-related behaviour over the first 6 months post-MI. However, their analyses did not control for pre-MI health-related behaviour. METHOD: This paper reports a re-analyses of the Auckland data, and a replication study conducted with 155 first-time MI patients in Brighton, United Kingdom (UK), to investigate whether baseline attributions for MI were related to health-related behaviour change at 6 months (N=132). Spouses (N=85) also completed the attribution questionnaire at baseline. RESULTS: There was no consistent relationship between the causal attributions of patients and subsequent behaviour change in Auckland and Brighton. For both samples, causal attributions were associated with pre-MI behaviour. CONCLUSIONS: The data from both samples suggest that the causal attributions of MI patients and their spouses may be realistic, but not predictive of subsequent changes in behaviour.