The multidimensional pathophysiology of septic shock is poorly understood and treatment modalities are controversial. The present study evaluates the relative importance of three therapeutic measures: antibiotics (trimethoprim and sulphamethovazole [TS]); fluid infusions (lactated Ringer's solution [RL] and 3% albumin [Alb]), and pharmacologic doses of corticosteroids (CS) (dexamethasone [DM]), using central hemodynamics (plasma volume [PV], cardiac output, oxygen consumption [&OV0312;O2]), and survival as end-points.Septic shock was induced by intraperitoneal injection of liveEscherichia colibacteria. At 5 h in untreated septic rats, PV had dropped to 76%, cardiac output to 69%, and &OV0312;O2to 71% of preshock levels. Untreated septic animals had a mean survival time of 9.7 ± 1.7 (SD) h, with none surviving 24 h. Regardless of therapy, cardiac output and &OV0312;O2at 10 h were predictors of survival time (p< .01). Treatment was initiated at 5.5 h after bacterial injection, at a time when TS therapy alone had not improved the 24-h survival rate. Animals treated with DM. RL, and Alb, in this order, exhibited progressively improved central hemodynamics, and 24-h survival rate increased to 60% compared with 0 in untreated animals (p< .001). The combination of DM and RL produced no further improvement. However, DM combined with 3% Alb restored &OV0312;O2cardiac output, and PV to 81%, 100%, and 125%, respectively, increasing the 24-h survival rate to 97% (29/30), significantly greater than that achieved by any other treatment modality (p< .05).