To measure specifically angiotensin-(1-8)octapeptide, peptides were extracted from 2 ml of plasma by reversible adsorption to bonded-phase silica. The angiotensin-(1-8)octapeptide was then isolated by isocratic reversed-phase high-performance liquid chromatography and quantified by radioimmunoassay. The extraction recovery of 125I-angiotensin II added to 2 ml of plasma was 99 +/- 2% (mean +/- SD). The overall recovery of 5, 10, and 20 fmol unlabeled angiotensin II added to 1 ml of plasma was 80 +/- 10%. The coefficient of variation for within-assay precision was 0.06 and for between-assay precision 0.13. The detection limit was 0.4 fmol/ml. Buffer and plasma blanks were below the detection limit. Normal subjects on a free diet in supine position averaged 4.2 +/- 1.7 fmol/ml angiotensin-(1-8)octapeptide. Furosemide (40 mg p.o.) and standing increased these values to 22 +/- 7.6 fmol/ml. In four volunteers, immunoreactive "angiotensin II" (more or less angiotensin-like material) was measured serially before and after converting-enzyme inhibition (Hoe 498) with conventional Dowex extraction. At peak inhibition, plasma immunoreactive "angiotensin II" levels decreased by only 44%. In contrast, angiotensin-(1-8)-octapeptide isolated by high-performance liquid chromatography completely disappeared. In hypertensive patients receiving long-term treatment with enalapril, plasma levels of angiotensin-(1-8)octapeptide fell from 2.7 +/- 0.9 to 0.9 +/- 0.3 fmol/ml (mean +/- SEM) 2 hours after the morning dose, whereas levels of immunoreactive "angiotensin II" were not significantly changed. We found that this sensitive method specifically measured angiotensin-(1-8)octapeptide and demonstrated that true angiotensin II virtually disappears during converting-enzyme inhibition.