In a series of 46 patients acute haemodilution was deliberately performed to test the validity of experimental data in clinical practice. Several metabolic and haemodynamic parameters as well as continuous registration of ECG were followed. The withdrawn blood has been exchanged by Dextran 60/PPL in equal quantities resp. by 5% human albumin. Using Dextran 60/PPL a light degree of hypervolemia (700 ml) was applied. Results confirmed that haemodilution is practicable in patients in principe. Exchange of blood by human albumin was followed by less severe side effect of haemodilution than was exchange by Dextran 60/PPL. in mild hypervolaemia. The course of haemodiluted and non haemodiluted patients during comparable surgical procedures (gastrointestinal operations) has been analysed: although measured parameters did not differ significantly between both series--except of Hb and Hk--side effects of the cardiovascular type were more frequently observed in the hemodilutiongroup. Provided there is knowledge of the dangers and side effects and special care (control monitoring) of the patients by the anesthesiologist, acute hemodilution is further recommended.