A correct estimation of volume status and so-called dry weight in dialysis patients remains a difficult clinical problem. Clinical status and chest X-ray are not sensitive enough, while invasively measured central venous pressures are not routinely available. Recently, the sonographic determination of the diameter and collapse of the inferior vena cava (IVC) has been proposed as a noninvasive method for estimating intravascular volume. We tried to evaluate the clinical relevance of this method in dialysis patients by comparing it with central venous pressures (CVP) and atrial natriuretic peptide (ANP). To establish a normal range and to control for confounding variables, we examined a large number of healthy controls. Furthermore, the influence of tricuspid insufficiency was examined echocardiographically. Measurements of the IVC diameters were well reproducible, with a coefficient of variation for interobserver error of 2.2%, and a coefficient of variation of 1.4% for intraobserver error. The collapse index was less well reproducible and therefore not used for further analysis. In 86 normal controls (age 18 to 76 years), IVC diameters showed a wide variation, and they were not correlated to age, height, weight, or body surface area. However, there was a significant correlation of IVCex to heart rate ( r =0.63, P r =0.72, P 12 cmH 2 O). In another 39 stable, chronic haemodialysis patients, there was a significant correlation of the intradialytic decrease of ANP and IVCex ( r =0.69, P per se . Intravascular volume changes are reflected by IVC measurements, as shown by the correlation to other indices of intravascular volume, such as CVP and α-hANP. IVC sonography is noninvasive and easily available; serial measurements allow an estimation of changes of intravascular volume in patients without cardiac filling impairment. However, unlike with body impedance, interstitial volume is not reflected by IVC diameters.