This article investigates patients' beliefs about the intervention offered to manage their illness. Coronary artery disease (CAD) patients, 70 of whom were undergoing medication, 71 to undergo angioplasty and 73 to undergo surgery, completed a 58-item questionnaire regarding their treatment beliefs. Responses were subject to principal components analysis, which indicated four factors accounting for 36.7 per cent of the variance. After excluding extraneous items, the final questionnaire consisted of 27 items, clustered around four components: treatment-value, treatment-concerns, decision-satisfaction and cure. A coherent set of subscale inter-correlations and ANCOVAs examining treatment group differences on these sub-scales showed a logical, explicable pattern of group differences reflecting the distinctive natures of each treatment and demonstrated discriminant validity. Correlations with other scales provided evidence of construct validity.