Clinical and hemodynamic effects of isovolemic hemodilution (HD) were evaluated in 12 patients (aged 59 +/- 8 years) with severe multivessel coronary artery disease (CAD) and angina pectoris grade III (Canadian Cardiovascular Society classification) despite high-dose medical treatment. In none of these patients was aortocoronary bypass grafting or percutaneous transluminal coronary angioplasty possible. Prior to HD and after 3 months of HD the incidence of angina pectoris was determined by means of questionnaires; hemodynamic measurements were performed with right heart catheterization at rest and during exercise. After 3 months of HD hematocrit was reduced from 46.2 +/- 1.3% to 38.5 +/- 0.5%. The weekly incidence of angina pectoris was unchanged (19 +/- 7 before, 17 +/- 8 after HD). Cardiac index was 2.5 +/- 0.7 1/min/m2 at rest and 3.9 +/- 1.0 1/min/m2 during exercise before, 2.6 +/- 0.5 1/min/m2 at rest and 3.9 +/- 0.8 1/min/m2 during exercise after HD. Stroke volume index did not increase significantly neither at rest nor during exercise after HD. Initially, systemic vascular resistance decreased from 1659 +/- 603 to 1398 +/- 420 dyns/cm5 during exercise; after HD it was 1522 +/- 551 (rest) and 1283 +/- 348 dyns/cm5 (exercise). Mean pulmonary artery pressure (PAP) and wedge pressure (WP) were unchanged at rest (PAP: 19.9 +/- 6.7 mm Hg before, 19.2 +/- 6.5 mm Hg after HD; WP: 10.8 +/- 5.5 mm Hg before, 10.7 +/- 4.3 mm Hg after HD) and during exercise (PAP: 43.0 +/- 9.9 mm Hg before, 42.8 +/- 8.9 mm Hg after HD; WP: 30.8 +/- 4.6 mm Hg before, 30.6 +/- 6.5 mm Hg after HD). In conclusion, in patients with CAD isovolemic HD does not reduce angina pectoris but also does not induce clinical deterioration. Furthermore, isovolemic HD does not worsen the hemodynamic effects of severe CAD with impaired left ventricular function.