Objectives: To assess, in clinical practice, the (i) incidence, (ii) within-patient consistency, and (iii) clinical spectrum of chest symptoms (chest symptoms) after subcutaneous (sc) and oral sumatriptan, and (iv) to identify risk factors for chest symptoms. Design: Two-year retrospective survey with mailed self-administered questionnaire. Setting: Neurolgy outpatient clinic of university hospital. Subjects: Migraine patients with or without aura (n=869). Main outcome measures: incidence, within-patient consistency and characteristics of chest symptoms; demographic and clinical characteristics of patients. Results: There were 735 (85%) respondents. Sumatriptan was used by 453 patients, during 25 months (median), for 28 000 attacks (median: 33 attackspatient). Of sumatriptan users, 41 % (sc) and 24% (oral) had chest symptoms in all attacks, 39% (sc) and 58% (oral) in none, and the remaining in some attacks. Because of chest symptoms, 10% discontinued sumatriptan. Chest symptoms mostly consisted of heavy arms and chest pressure, started within 5 (sc) to 30 (oral) min, and lasted 30 (sc) to 60 (oral) min. Compared with patients without chest symptoms, patients with chest symptoms more often were females and younger, and went to rest immediately after sumatriptan administration (alI p<0.001); they also tended to have lower body mass indices, more severe attacks and less efficacy of sumatriptan (all 0.001<p<0.05). Patients with chest symptoms had no higher incidence of cardiovascular symptoms or risk factors. Conclusions: Chest symptoms are frequent, within-patients consistent, but rarely important, adverse events of (notably sc) sumatriptan. The risk of chest symptoms is patient-dependent and not related, even opposite, to cardiovascular disease. This contradicts the hypothesis that chest symptoms after sumatriptan are caused by cardiac ischemia. Patient acceptance of chest symptoms is improved by pre-advising of the risk and nature of chest symptoms.