Extracellular volume can be estimated from the distribution volume of sucrose (Vdsucrose). The purpose of this study was to establish sucrose pharmacokinetics in preterm infants less than 1500 g compared to children and adults and to define an optimal sampling scheme. In five preterm infants, 10 children, and five adults Vdsucrose after a single injection was calculated with the two-compartment model (Vdsucrose-TCM) and with the one-compartment model applied only to the elimination phase of the same concentration-time curve (Vdsucrose-OCM). In preterm infants Vdsucrose-TCM was 417 +/- 45 mL/kg (mean +/- SD). Vdsucrose-OCM was only 3.0 +/- 2.3% higher, because sucrose elimination half-life was on average 250 times longer than distribution half-life. Therefore Vdsucrose-OCM, requiring only four blood samples between 2 to 5 h after injection, gave an adequate estimate of Vdsucrose in preterm infants less than 1500 g. Vdsucrose-TCM in children and adults was 188 +/- 26 and 189 +/- 17 mL/kg, respectively. Vdsucrose-OCM was 10 to 65% higher. Therefore, in children and adults only Vdsucrose-TCM gives a reliable estimate of Vdsucrose. This requires 10 to 15 blood samples. The reduced sampling scheme was used in an extension of the study of preterm infants including five additional infants. Vdsucrose-OCM in the preterm infants was 462 +/- 47 mL/kg at birth and 425 +/- 46 mL/kg at maximal postnatal wt loss. Postnatal wt loss (mean -83 +/- 44 g) was not significantly different from postnatal reduction of Vdsucrose-OCM (mean -82 +/- 56 mL), suggesting that postnatal wt loss mainly represents extracellular fluid loss.