We looked for the presence of prorenin in erythrocytes from normal subjects (n = 8), hypertensive patients (n = 8), and pregnant women (n = 8). Angiotensin I generation was measured at 37 degrees C, pH 5.7, in the presence of homologous substrate (1400 ng/mL) before and after trypsin activation (100 micrograms/mL) in (A) haemolyzed erythrocytes, (B) supernatants of haemolyzed erythrocytes, and (C) in the sixth washing of erythrocytes diluted 1:1 with a 0.1 M Tris buffer containing 0.5% bovine serum albumin and protease inhibitors. Haemolyzed erythrocytes generated angiotensin I only after trypsin treatment, and the rate of generation was the same (A) before and (B) after centrifugation at 20,000g, indicating the absence of prorenin bound to the cell membranes. When aliquots of the last washing of erythrocytes (C) were tested for angiotensin I generation before and after trypsin, they did not generate angiotensin I, indicating that residual prorenin from the plasma was no longer present in our preparation. Angiotensin I generation by trypsin-treated A and B was completely abolished by preincubation with anti-renin serum. The level of prorenin was not significantly different in the erythrocytes from normal, hypertensive, and pregnant subjects (68 +/- 10, 58 +/- 7 and 107 +/- 17 pg angiotensin I.mL-1.h-1, ns) in spite of their very different plasma levels (21 +/- 2.5, 17 +/- 2.4 and 110 +/- 12 ng angiotensin I.mL-1.h-1, p less than 0.01 for pregnant women compared with both normal and hypertensive subjects). Our data show that prorenin is present in human erythrocytes in fairly constant and clearly detectable amounts, thus suggesting a possible intracellular role for it.