The mortality pattern among 999 Danish patients who had been subjected to angiography of the carotid artery with the alpha-ray emitting X-ray contrast media Thorotrast during the period 1935-47 was assessed by record linkage with the National Death Registry through 1989. Standardized mortality ratios (SMR) were calculated relative to the general population. The overall SMR was increased by 18 times during the first 3 years after Thorotrast injection. This rate reflects the often serious, underlying neurological conditions for which angiography was performed, however, mortality was increased by 3-4 fold even for the follow up period after the first 3 years. The increase in mortality was evident for all categories of cause of death, the SMR being 11.1 (95% confidence interval (CI) 7.1-16.4) for cirrhosis of the liver, 4.7 (4.1-5.3) for cancer, 1.6 (1.3-1.9) for cardiac disease, 3.3 (2.6-4.2) for cerebrovascular diseases, 3.9 (3.3-4.5) for other natural causes, and 4.4 (3.4-5.6) for violent causes (including suicides). The SMR was generally related positively to young age at injection, to time since injection, and to the amount injected. The excess mortality can be explained only partially by underlying neurological conditions and by diseases known to be induced by Thorotrast (cirrhosis and cancer of the liver, leukaemia and other haematological diseases), and it is suggested that unspecific effects induced by the alpha-radiation of Thorotrast may have contributed.