The recent identification of messenger RNAs encoding renin and angiotensinogen in nonrenal tissues raises the possibility that angiotensins (Ang) may be formed extrarenally and released into the plasma. The aim of this investigation was to test the hypothesis that plasma angiotensins may originate from extrarenal sites. Twenty-five patients with chronic renal failure (six surgically anephric and 19 with kidneys in situ) were studied prior to and after a standard hemodialysis treatment. Angiotensins were measured by extraction. high-pressure liquid chromatography (HPLC) separation, and radioimmunoassays. In patients with kidneys present, plasma renin activity (PRA) was 3.1 ± 0.7 ng Ang I/ml/h. Ang I, Ang II. and Ang III levels were 70.6 ± 9.0. 44.0 ± 9.8, and 20.2 ± 3.6 pg/ml, respectively. In all six anephric patients PRA was undetectable ±0.1 ng Ang I/ml/h). Ang I and Ang II were detected in four anephric patients, and Ang III was detected in three anephric patients (Ang I, (10.4 ± 5.2; Ang II, 2.6 ± 1.2; Ang III. 2.7 ± 1.5 pg/ml, n = 6). At the completion of dialysis treatments, which reduced body weight by 2.5 ± 0.2 kg in patients with kidneys and by 2.1 ± 0.3 kg in anephric patients, there were no significant changes in PRA or plasma angiotensins in either group. Reduction in body water by hemodialysis did not increase the concentration of angiotensins in plasma. We conclude that there is a small but definite component of plasma angiotensin that is produced by nonrenal mechanisms and that is not stimulated by volume depletion.