Between 1982 and 1988, we observed 312 patients who were affected by syncope or presyncope and whose spontaneous symptoms could be reproduced by means of carotid sinus massage (CSH); no other definite cause of syncope could be identified. The clinical outcome during a 2- to 8-year follow-up period (mean 44 +/- 24 months) was assessed in 262 of them (mean age, 71 +/- 11 years; 183 men) and was compared with that of a group of 55 patients who were affected by unexplained syncope (control patients) who were matched 4:1 for age and sex with CSH patients. CSH patients had an overall mortality rate of 7.3 per 100 person-years (cardiovascular, 66%; sudden death, 9%); overall predicted cumulative survival rates at 1, 3, 5, and 7 years were 92%, 80%, 66% and 53%. Survival was similar in control patients; mortality rate was 5.8 per 100 person-years (cardiovascular, 82%; sudden death, 18%); cumulative survival rates at 1, 3, 5, and 7 years were 85%, 80%, 73%, and 69%. Standardized mortality rate of the general population with similar age and sex distribution, as calculated by means of Italian Istituto Centrale di Statistica death-rate data (1987 edition) was 8 per 100 person-years. Of 13 clinical variables, age, sex, abnormal electrocardiogram, and heart failure (but not CSH type or related arrhythmias) were independently linked to mortality in CSH patients (Cox model). In the vasodepressor form of CSH, patients were younger than those with other forms of CSH and the percentage of women was higher.(ABSTRACT TRUNCATED AT 250 WORDS)