Factors affecting cardiac function in dialysis patients include arterial blood pressure, anaemia, intravascular volume, and the arteriovenous (a-v) access. Cardiac failure has been directly attributed to dialysis a-v access in several cases. The contribution of the a-v access to cardiac performance has been tested, in the past, by a short manual compression on the fistula, but this technique has obvious limitations. The present study examined prospectively the effect of dialysis a-v access on both cardiac function and various hormonal responses. Ten patients (age, mean +/- SD, 59.6 +/- 12.3) with end-stage renal failure being prepared for chronic dialysis therapy were included. All patients underwent an echocardiographic study before and 2 weeks after the creation of the a-v access. Plasma atrial natriuretic peptide (ANP), plasma renin activity (PRA), and plasma aldosterone were measured at the same time periods. Following the creation of the a-v fistula or graft, shortening fraction increased by 15.8 +/- 6.3% (P < 0.01), stroke volume increased by 21.9 +/- 5.3% (P < 0.01), ejection fraction increased by 10.6 +/- 4.5% (P < 0.02), cardiac output increased by 19.0 +/- 6.9% (P < 0.02), and cardiac index increased by 18.3 +/- 7.1% (P = 0.05). Systemic vascular resistance decreased by 23.5 +/- 7.1% (P < 0.01). There was no change in blood pressure, heart rate, weight, haemoglobin or serum creatinine. ANP increased by 83.7 +/- 17.0% following the a-v access operation (P < 0.001), PRA decreased by 41.2 +/- 10.0% (P < 0.05), and plasma aldosterone did not change. None of the patients developed overt high-output cardiac failure. This study shows that at least in the short term following the creation of a dialysis a-v access, a mild state of volume overload develops, which is offset by the ¿unloading' effect of the decreased peripheral vascular resistance; the latter is probably mediated by secretion of ANP in response to atrial stretching.