To study the effect of oncotic pressure on brain water content during cardiopulmonary bypass (CPB), 14 anesthetized New Zealand White rabbits underwent 60 min of nonpulsatile CPB at normothermia. Animals were grouped according to the composition of the circuit priming fluid. Group 1 animals (n = 7) received a priming fluid (6.5% hydroxyethyl starch in 0.72 N NaCl; 323 +/- 13 mOsm/kg [mean +/- SD]) that maintained normal colloid oncotic pressure (COP) during CPB (19.0 +/- 1.5 mmHg). Group 2 animals (n = 7) received a priming fluid (0.9 N NaCl; 324 +/- 23 mOsm/kg) that led to a hypooncotic state (COP = 6.2 +/- 1.2 mmHg). Blood chemistries and hemodynamics were recorded every 15 min during CPB. Animals were given additional priming fluid and sodium bicarbonate during CPB to maintain a circuit flow of 85 ml.kg-1.min-1 and arterial pH greater than 7.35. There were no significant differences between groups 1 and 2 with respect to temperature, central venous pressure, mean arterial pressure, PaO2, PaCO2, plasma sodium concentration, or osmolality at any time during CPB, although osmolality increased in both groups. After 60 min of bypass, animals were killed and organ water contents were determined by wet/dry weight ratios. A separate group of nine similarly prepared and anesthetized animals that did not undergo cannulation or CPB also underwent measurement of plasma chemistries and tissue water contents and served as nonbypass controls (group 3). Brain and kidney water contents were unaffected by oncotic pressure, whereas duodenum and skeletal muscle had significantly greater water content (P = 0.003 and P = 0.008, respectively) after hypooncotic CPB.(ABSTRACT TRUNCATED AT 250 WORDS)