Coronary artery surgery is usually successful in relieving angina but benefits for quality of everyday life are sometimes disappointing. Generalized linear inter-active modelling (GLIM) was used to examine pre-operation predictions of psychological and social outcome 3 and 12 months after coronary artery surgery. The study identified predictors of return to work (social class, previous employment status), psychiatric outcome (pre-operation mental state) and social adjustment (pre-operation mental state, age, social class). Although our analysis of potential predictors of 1 yr outcome at 3 months after surgery was less detailed, early convalescence is likely to be the best time for clinical identification of those at risk of poor long term outcome. The findings have implications for selection and preparation for surgery and identification of those who might benefit from extra rehabilitation.