The value of the tilt test in assessing unexplained syncope in children and adolescents was investigated in 79 patients (41 girls and 38 boys, 12.3 +/- 2.9 years). The test was performed with a tilt of 60 degrees for 45 minutes. A pharmacological provocation with isoprenaline was carried out in negative passive test. The tilt test was considered to be positive in cases of symptomatic falls of the systolic blood pressure with or without bradycardia. Asystole was defined as a ventricular pause > or = 5 seconds. During the investigation, 52 patients (66%) reproduced symptoms related to vasovagal syncope in 45 patients (57%) or panic attack in 7 others (9%). Of the 45 presenting vasovagal syncope, 7 had a cardio-inhibitory reaction with asystole of 11.2 +/- 3.3 seconds. The other responses were vasodepressive in 17 patients and mixed in 21 patients. A comparative study did not reveal any significant difference between patients with a positive or negative test including those with panic attacks. Seventy-five patients were followed up for an average of 32 +/- 22.3 months. Preventive therapy was prescribed in 34 patients because of the frequency or severity of their symptoms. At the end, 10 patients (13%) had at least one recurrence of syncope, 3 in the group with panic attacks, 5 in non-treated patients (irrespective of the result of the tilt test) and only 1 in the group of 23 patients treated with beta-blockers. The authors conclude that with a diagnostic return of 66% in this series, the tilt test seems to be the non-invasive investigation of choice in unexplained syncope in children and adolescents. In addition, true vasovagal syncope can be differentiated from psychogenic.