Secretion of ANP (atrial natriuretic peptide) and BNP (brain natriuretic peptide) is pulsatile in healthy humans. However, the patterns of secretion of ANP and BNP have not been studied in chronic renal failure. The aim of the present study was to test the hypotheses that ANP and BNP are secreted in pulses in dialysis patients, and that pulsatile secretion is regulated by prostaglandins. Blood samples were drawn every 2 min through an intravenously inserted plastic needle over a period of 1-2 h in 13 dialysis patients and 13 healthy control subjects (Study 1), and in 15 healthy control subjects, who participated in a randomized placebo-controlled cross-over study after treatment with indomethacin and placebo (Study 2). Plasma concentrations of ANP and BNP were determined by RIAs, and the results were analysed for pulsatile behaviour by Fourier transformation. The results from Study 1 showed that the secretion of ANP and BNP was pulsatile in nine patients with chronic renal failure. The maximum amplitude was significantly higher in chronic renal failure compared with control subjects for both ANP and BNP (ANP, 4.3 compared with 0.7 pmol/l; BNP, 2.0 compared with 0.3 pmol/l; values are medians) and correlated positively with the mean plasma level of ANP (rho=0.900, P=0.001; n=9) and BNP (rho=0.983, P=0.000; n=9). The frequency was the same for patients and controls. The results from Study 2 demonstrated pulsatile secretion in all subjects, but both the amplitude and frequency were unaffected by indomethacin. The maximum amplitude correlated positively with the mean plasma level of ANP and BNP during both placebo and indomethacin treatment. It can be concluded that the secretion of ANP and BNP is pulsatile with abnormally high amplitude in chronic renal failure, that prostaglandins apparently are not involved in the secretion of these peptides in healthy subjects and that the high secretion rate in chronic renal failure results in higher ANP and BNP in plasma.