Prolonged 60 degree head-up tilt has been shown to be valuable in the investigation of unexplained syncope, diagnosing neurally mediated bradycardia/hypotension or malignant vasovagal syndrome. To evaluate the methodology of tilt testing, the following were examined: reproducibility of results, tilt duration, angle of tilt, method of tilt support and effect of age in patients and control subjects. Seventy-one patients with recurrent unexplained syncope underwent 60 min of 60 degree tilt; 53 (75%) had an abnormal test with vasovagal syncope at 24 +/- 10 min (mean +/- SD). Tilting to 60 degrees resulted in an abnormal test in only 2 (7%) of 27 control subjects without cardiovascular symptoms (p less than 0.001); and 5 (15%) of 34 patients with syncope and documented conduction tissue disease (p less than 0.001). Of 15 youthful fainters, 3 (20%) had vasovagal reactions as did 1 (8%) of 12 asymptomatic youthful control subjects. These 12 control subjects also underwent tilting with a saddle support and 7 (67%) had vasovagal reactions. It is concluded that the duration of tilting at 60 degrees should be 45 min (mean time to syncope +2 x SD in the 53 patients with abnormal results). Twenty percent of patients with an abnormal tilt test may not demonstrate syncope with repeat tilting. Saddle tilt testing in unexplained syncope may result in a loss of specificity. Tilting at less than 60 degrees results in a loss of sensitivity. Head-up tilt may be less useful in youthful subjects with vasovagal syncope than in other subjects.