Four patients with rheumatoid constrictive pericarditis and two patients with rheumatoid cardiac tamponade are presented, and 60 previously reported cases with these two complications are reviewed. Rheumatoid arthritis was moderate to severe in 84% of the patients with cardiac tamponade and in 74% of the patients with constrictive pericarditis. However, both these complications were also seen in patients who had only mild arthritis and in two previously reported cases constrictive pericarditis actually preceded the onset of rheumatoid arthritis. The duration of rheumatoid arthritis had no bearing on the development of these complications. In 75% of patients with cardiac tamponade, and in 66% of cases with constrictive pericarditis, subcutaneous nodules were present. In those cases where the rheumatoid factor was measured it was positive in 92% with cardiac tamponade and in 84% with constrictive pericarditis. In 63% of patients with cardiac tamponade and in 70% of cases with constrictive pericarditis a history of pericardial type of pain was obtained and/or a pericardial rub heard. The diagnosis of cardiac tamponade and constrictive pericarditis was made clinically and in doubtful cases confirmed by cardiac screening and intracardiac pressure recordings. The low sugar content in the pericardial fluid in the absence of infection or malignancy was an important clue to the rheumatoid etiology of the effusion. In the majority of the cases histological appearances of the pericardial tissue showed non-specific fibrous reaction and infiltration with plasma cells and lymphocytes. Only in five of the cases, including one from the present series, were typical rheumatoid granulomatous lesions demonstrated. Treatment with corticosteroids neither prevented the occurrence nor led to amelioration of either cardiac constriction or tamponade. Pericardial resection was life saving, producing both symptomatic and objective involvement of the cardiac function. In the present series of six cases two patients developed aortic incompetence. In one of these it was due to rheumatoid granulomatous valve disease and in the other due to non-specific aortic valvulitis. The combination of constrictive pericarditis and granulomatous aortic valve disease has not been previously recorded.