The colloid osmotic pressure (COP) is not routinely assessed during pediatric heart surgery. Two cases of unrecognized hyperoncotic states associated with renal failure have been observed after pediatric heart surgery. We studied the hypothesis that the COP cannot be estimated from the total plasma protein (TPP) or albumin level. The course of COP and its correlation to the TPP and albumin level were investigated in 25 children undergoing elective heart surgery. Infusion therapy was performed solely on the basis of clinical parameters and TPP/albumin levels. COP values were determined in a blinded fashion at the end of the study. No correlation between TPP/albumin and the COP could be determined preoperatively. On arrival at the ICU correlation was strong. A weak correlation was observed at 24 hours and 48 hours after surgery. However, the observed wide range of the confidential bands indicates that the COP cannot be estimated correctly, neither from the TPP, nor from the albumin level. Due to colloidal oversubstitution COP was significantly increased compared to preoperative level at 48 hrs following surgery. As estimation of COP from TPP or albumin level is inaccurate, oncometry should be performed during pediatric heart surgery.