Recent evidence suggests that a considerable proportion of plasma angiotensin is generated not in blood but in peripheral tissues. Through the measurement of angiotensin peptides and renin in the plasma of 11 anephric subjects, we have investigated whether kidney-derived renin, or some other tissue mechanism for angiotensin generation, is the major determinant of plasma angiotensin. Particular care was taken to prevent inadvertent activation of inactive renin and possible generation, conversion and metabolism of angiotensin peptides during processing of blood samples. Initial experiments revealed that plasma from anephric subjects contains high amounts of material which interferes in radioimmunoassays for angiotensin, even after high-performance liquid chromatography (HPLC). Therefore, in order to obtain an unambiguous identification of angiotensin peptides, a dual HPLC method was developed in which angiotensin peptides were first separated by HPLC, then acetylated and run again on HPLC before radioimmunoassay for angiotensin I and II (detection limits, 0.25 and 0.2 fmol/ml, respectively). The levels of angiotensin I and II were 1.2 +/- 1.6 and 0.7 +/- 0.5 fmol/ml (mean +/- s.d., n = 9-10), respectively, being 6% of levels in normal subjects, and were consistent with the active renin levels (1.8 +/- 1.7 muIU/ml, n = 11) which were 7% of levels in normal subjects. Artefactual activation of prorenin and angiotensin generation during sample processing were excluded as significant causes of the low levels of active renin and angiotensin I and II in anephric plasma. These data indicate that kidney-derived renin is the major determinant of angiotensin levels in normal human plasma. However, the present demonstration of low levels of active renin and angiotensin I and II in plasma of anephric subjects provides unequivocal evidence for a functional extrarenal renin-angiotensin system in man.