Three series of experiments were conducted to develop a model of volume-controlled severe hemorrhagic shock in the unanesthetized analgesic cynomolgus monkey. This report concerns the insult without resuscitation. In Series I, seven monkeys were sedated with 75% N2O/25% O2, bled 40% of their measured blood volume over 20 min and observed until death. Mean arterial pressure (MAP) decreased to 21 +/- 6 mmHg, spontaneously increased to 46 +/- 5 mmHg, then gradually decreased to pulselessness at 146 +/- 42 min (range 101-213). Hemodynamic variables, lactate, base excess, electroencephalogram and sagittal sinus PO2 followed the same biphasic pattern. In Series II, eight monkeys were bled 27 ml/kg (43% of estimated blood volume) over 20 min under the same N2O analgesia and with similar responses as in Series I. In Series III 26 monkeys were bled 27 ml/kg over 20 min (time zero) as in Series II. Three developed apnea and pulselessness at end of hemorrhage. In 23 the shock period was prolonged for testing resuscitation therapies. Starting at 0 + 30 min, MAP was controlled with minute blood volume adjustments at 30 mmHg until 0 + 2 h. Three died due to inaccurate (preventable) MAP adjustments. At MAP 30 mmHg, all animals lost consciousness, EEG activity decreased, and brain stem reflexes disappeared. The "volume-pressure controlled" hemorrhagic shock model of Series III retains the initial natural response to bleeding, simulates the clinical picture of severe prolonged shock without anesthesia, and represents a more controllable insult than volume controlled hemorrhage alone.