Cardiac involvement, assessed yearly by a semiquantitative cardiac scoring technique (the sum of 2 variables, scored 0 or 1 for left axis deviation, and 0 or 2 for moderate-large pericardial effusion) and outcome at 10 years were evaluated in 90 systemic sclerosis (SSc) patients enrolled in a 3-year prospective drug trial. Higher cardiac scores were associated with decreased survival. The inverse relationship of survival with cardiac score was more significant than the relationship between survival and any of the cardiopulmonary variables individually (P less than 0.0000003). Predicted 6-year survival was 79% in patients with a cardiac score of 0, 51% in those with a score of 1, 15% in those with a score of 2, and 0% in those with a score of 3. We conclude that a simple 2-variable weighted cardiac score, derived from routine electrocardiography and echocardiography, is a useful predictor of survival in patients with SSc, and we propose that the application of this cardiac score be tested in other populations of SSc patients.