In patients with cirrhosis a diminished effective central arterial blood volume associated with systemic arterial vasodilatation has been proposed as the mechanism that initiates renal sodium retention. Furthermore, total central blood volume has recently been reported as reduced in cirrhosis, and the controversy over the stimulus for sodium retention in cirrhosis remains. The aim of this study was to assess the central blood volume with radionuclide angiography to determine whether there is effective arterial underfilling in cirrhosis. Twenty-nine patients (13 with and 16 without ascites) and 10 age- and sex-matched control subjects were studied under metabolic conditions. Radionuclide ventricular volume and total central blood volume were determined from gated images, taking into account the 99Tc count activity per milliliter of blood volume and attenuation. The pulmonary volumes were similarly derived. The cirrhotic patients as a group had significantly higher total central blood volume (1,287 ± 105 ml/m2 in control subjects vs. 1,874 ± 106 ml/m2 in cirrhotic patients, p < 0.01), right and left pulmonary blood volumes (217 ± 20 ml/m2 in control subjects vs. 309 ± 20 ml/m2 in cirrhotic patients, p = 0.03 and 185 ± 18 ml/m2 in control subjects vs 288 ± 22 ml/m2 in cirrhotic patients, p = 0.02, respectively), cardiac and central vascular blood volume (885 ± 79 ml/m2 in control subjects vs. 1,276 ± 75 ml/m2 in cirrhotic patients, p = 0.01), cardiac output (5.36 ± 0.56 L/min in control subjects vs. 7.19 ± 0.50 L/min in cirrhotic patients, p = 0.05), heart rate (65 ± 3 beats/min in control subjects vs. 75 ± 2 beats/min in cirrhotic patients, p = 0.04) and significantly lower systemic vascular resistance (1,443 ± 121 dyne·sec·cm−5 in control subjects vs. 1084 ± 68 dyne·sec·cm−5 in cirrhotic patients, p = 0.02). All volumes were significantly higher in both the nonascitic and the ascitic patients when compared with the control subjects. Significantly increased cardiac output and reduced systemic vascular resistance, however, were only observed in the ascitic patients. Neurohumoral markers were increased in the ascitic patients compared with both the control subjects and nonascitic patients, but the difference was not statistically significant. There was no correlation between any of the volume measurements with neurohumoral markers of an effective arterial blood volume. In conclusion, the ascitic patients with cirrhosis have evidence of central hypervolemia and hyperdynamic circulation, as indicated by peripheral vasodilatation, an increased cardiac output and heart rate. In contrast, the nonascitic patients with cirrhosis who have previously been shown to have a compensated sodium handling abnormality demonstrate an increased total central blood volume without significant peripheral vasodilatation and evidence of arterial underfilling. This suggests that peripheral vasodilatation is not solely responsible for the sodium retention and central blood volume expansion in cirrhosis. (Hepatology 1994;19:312–321).