In 88 of 103 consecutive patients with angina and normal coronary arteries, follow-up data could be achieved 6-11 years (9.2 +/- 1.2 years) after diagnostic left heart catheterization. Three of these patients died during follow-up (two noncardiac deaths and one death with no identifiable etiology). One patient suffered a documented myocardial infarction. In 40 patients (47%) chest pain diminished, while symptoms were unchanged in 20 (24%) or even worse in 25 (29%). Resolution or persistence of chest pain could not be predicted either by the character of pain (typical vs. atypical), the presence of hypertension, a left bundle-branch block, a positive exercise electrocardiogram or pathological pulmonary artery pressures during exercise, documentation of myocardial bridges, local wall motion abnormalities, or a left ventricular end-diastolic pressure greater than or equal to 13 mmHg. However, continuing chest pain was significantly more common in patients who revealed a 'slow-flow phenomenon' at initial coronary arteriogram. Thus, in patients with angina and normal coronary arteries the long-term course regarding frequency of morbid cardiac events is benign. However, more than half of the patients reported chest pain to be similar or even worse than at catheterization. Most clinical and invasive results at initial evaluation had no predictive value for the persistence of symptoms. The impact of 'slow-flow' in coronary arteriography, which was a phenomenon almost exclusive to patients with constant or even worse chest pain at follow-up, should be evaluated in larger patient populations.