An experimental design including an external closed-loop PID-(proportional-integral-differential-)controller is presented which enables the induction of gradual hemorrhagic hypotension at different stages of blood flow reduction up to stages of critically disturbed systemic and regional hemodynamics and oxygen supply. For this purpose nine newborn piglets (12-26 hours old, body weight 1626+/-160 g) were anesthetized and artificially ventilated. Gradual hemorrhagic hypotension was induced at four different steady state stages (stage 1 = 60 mmHg; stage 2 = 50 mmHg; stage 3 = 40 mmHg; stage 4 = 35 mmHg) every 30 minutes by gradual blood withdrawal using external PID controller equipment. Cardiac output and brain regional blood flows were measured by the colored microsphere technique. Systemic and brain regional hemodynamics and O2 supply, metabolic parameters and blood catecholamine concentrations were obtained under baseline conditions and at every 25th minute of the four different steady state stages. About 35 percent of the calculated total blood volume (cTBV) was withdrawn in order to reach the first stage of hemorrhagic hypotension. Further blood withdrawal of about 10 percent of the cTBV, about 5 percent of the cTBV, and about 3 percent of the cTBV were necessary to reach the other respective hypotensive stages. Gradual hemorrhagic hypotension led to an increasing reduction of the cardiac output at every hypotensive stage up to about 20 percent of the baseline value (p<0.05). This was accompanied by a concomitant increase of the total peripheral resistance to about 2.5 fold (p<0.05) and a huge increase in the blood catecholamine concentrations (epinephrine: about 64 fold; norepinephrine: about 35 fold). The induced redistribution of the circulating blood volume was shunted to the vital organs. Therefore, brain cortical blood flow was slightly increased at stage 1 and stage 2. A significant reduction of rCBF did not occur until stage 4 (p<0.05). Regional cerebrovascular resistance was concomitantly reduced at stage 1 and stage 2 (p<0.05) and thereafter again slightly elevated. Brain cortical oxygen consumption was maintained up to stage 2, reduced by about 20% at the next stage of hemorrhagic hypotension (p<0.05) and reached the lowest level of about 50% from baseline at stage 4 (p<0.05). Excellent accuracy and stability was shown at each stage for the external PID controller equipment, so that each given setpoint of the instantaneous mean arterial blood pressure was reached and stabilized even at the lowest hypotensive stage (stage 1: 59.53+/-0.23; stage 2: 50.03+/-0.56; stage 3: 39.18+/-1.75; stage 4: 35.28+/-0.45 mmHg (mean+/-SD)). We conclude that the experimental design presented, with an external PID controller to induce gradual hemorrhagic hypotension in newborn piglets is sufficient for producing functional states with changed systemic and cerebral features with high stability and accuracy, enabling a systematic study of disturbed regional hemodynamics and energy metabolism under steady state conditions even under critically changed states of the systemic cardiovascular regulation.