A retrospective analysis was undertaken of all surgical patients in Malmö, Sweden, during the period from 1951 to 1988 in whom pulmonary emboli were found at autopsy. The analysis included a comparison with the earlier analysed and reported time period from 1951 to 1980. A continued high frequency of pulmonary embolism at autopsy was seen during the 1980s (20.3 per cent of deaths, 31.7 per cent of autopsies). Of the 391 autopsy-verified pulmonary emboli found between 1981 and 1988, 113 (28.9 per cent) were considered fatal, 104 (26.6 per cent) contributed to death and 174 (44.5 per cent) were incidental. Few patients had symptomatic deep vein thrombosis or pulmonary embolism before death. The overall frequency of major pulmonary embolism in surgical patients remained unchanged (0.3 per cent). The frequency of major postoperative pulmonary embolism showed an increase during the 1950s and 1960s (maximum 0.4 per cent) but a decrease in the last 5-year period of the 1970s (0.3 per cent), which has earlier been reported upon. This decrease continued during the 1980s (0.2 per cent) (P less than 0.05). An increase was found in the number of patients operated on who had autopsy-proven pulmonary embolism and who received thromboprophylaxis. Pulmonary embolism continues to be a major cause of death in surgical patients; however, postoperative major pulmonary embolism has shown a reduction in the last 15-year period.