Different animal models have been used to study the relationship between oxygen consumption (VO2) and oxygen delivery (DO2) and each of them has its own specificity. The present study compared the models of balloon inflation in the inferior vena cava (BALL), tamponade (TAMP), and hemorrhage (HEM) to acutely reduce blood flow in pentobarbital-anesthetized dogs. The order of the procedures was randomized first, but HEM was irreversible so that HEM was applied as the third procedure (n = 10) or as a unique procedure (n = 4). Critical DO2 (DO2 crit) was determined as the intercept of two best fitted lines of regression of VO2 against DO2, but similar results were obtained when it was calculated from blood lactate or from arterio-venous differences in pH (AVpH) or PCO2 (VAPCO2). At DO2 crit, VO2 was 6.6 +/- .7 mL/min.kg in BALL, 6.2 +/- .6 mL/min.kg in TAMP and 6.0 +/- .7 mL/min.kg in HEM (p < .05 vs. BALL). DO2 crit was significantly greater with BALL (14.3 +/- 2.3 mL/min.kg) than with TAMP (10.2 +/- 1.7 mL/min.kg, p < .01 vs. BALL) or HEM (9.1 +/- 1.4 mL/min.kg, p < .01 vs. BALL). Also, critical oxygen extraction (EO2 crit) was significantly lower with BALL (46.5 +/- 4.8%) than with TAMP (62.1 +/- 11.5%, p < .01 vs. BALL) or HEM (66.8 +/- 11.4%, p < .01 vs. BALL, p < .05 vs. TAMP). The earlier onset of tissue hypoxia with BALL was probably related to the blood flow redistribution induced by this model.(ABSTRACT TRUNCATED AT 250 WORDS)