Ten percent pentastarch is a low-molecular-weight hydroxyethyl starch with greater oncotic pressure and shorter intravascular persistence than 6% hetastarch. To evaluate its safety and efficacy as a component of cardiopulmonary bypass priming solution, we prospectively studied 90 patients undergoing coronary artery bypass grafting or valve replacement necessitating cardiopulmonary bypass (bubble oxygenator and moderate systemic hypothermia). Sixty patients were randomized to receive 75 gm of either 10% pentastarch (group P) or 25% albumin (group A), and 30 patients received lactated Ringer's solution alone (group C). Intravascular colloid osmotic pressure during cardiopulmonary bypass was highest with either of the colloid primes (15-minute measurement: group P, 15.7 +/- 2.2 mm Hg (mean +/- standard deviation); group A, 15.2 +/- 2.0 mm Hg; group C, 11.3 +/- 1.7 mm Hg; p less than 0.05, groups P and A compared with group C). This was associated with a lower volume requirement during cardiopulmonary bypass to maintain the venous reservoir (group P, 333 +/- 318 ml; group A, 483 +/- 472 ml; group C, 1332 +/- 1013 ml; p less than 0.05, groups P and A compared with group C). Urine output during cardiopulmonary bypass was similar in each group. Net intraoperative fluid balance was lowest in the colloid groups (groups P and A, 5.7 +/- 1.4 L; group C, 6.9 +/- 1.3 L; p less than 0.05, groups P and A compared with group C). Cardiac index shortly after weaning from cardiopulmonary bypass was greatest in group P (group P, 3.2 +/- 0.9; group A, 2.8 +/- 0.8; group C, 2.7 +/- 0.6 dyne.sec.cm-5; p less than 0.05, group P compared with group C). Changes in alveolar-arterial oxygen gradients, shunt fraction, and effective compliance were similar in all groups. During cardiopulmonary bypass, pentastarch appeared to cause the greatest degree of hemodilution, as suggested by the lowest hemoglobin, factor VII and IX levels and platelet count. The activated partial thromboplastin time was significantly prolonged during and immediately after cardiopulmonary bypass in group P relative to groups A and C (p less than 0.05), although there were no significant differences in the activated clotting time before cardiopulmonary bypass, during cardiopulmonary bypass, or after heparin neutralization. As well, clinical indices of hemostasis, including mediastinal drainage, red cell, platelet, and fresh frozen plasma requirements, and reoperation for excessive postoperative bleeding, were similar. We conclude that pentastarch, when used in cardiopulmonary bypass prime, is as safe as either albumin or Ringer's solution alone.(ABSTRACT TRUNCATED AT 400 WORDS)