Objective: To investigate the relationship of mortality to early resuscitation using two levels of oxygen delivery ((D) over dot o(2)) in critically ill surgical patients greater than or equal to 50 yrs of age who were stratified into groups: age less than or equal to 75 yrs (age 50 to 75 yrs group); and age >75 yrs (age >75 yrs group), Design: A prospective, randomized trial, continued from a previous project, Setting: Surgical intensive care unit, university affiliated, Patients: Consecutive patients,greater than or equal to 50 yrs of age, unable to generate a (D) over dot o(2) of >600 mL/min/m(2) with fluid resuscitation alone, with a diagnosis of systemic inflammatory response syndrome, sepsis, severe sepsis, septic shock, and/or acute respiratory distress syndrome, Interventions: During the first 24 hrs of resuscitation, patients were randomized to receive fluids, blood transfusions, and vasoactive agents in order to achieve bo, treatment seals of >600 mL/min/m(2) in the protocol group and 450 to 550 mL/min/m(2) in the control group, Measurements and Main Results: One hundred five patients completed the study. In patients aged 50 to 75 yrs, the mortality rate was 21% (9/43) in the protocol group and 52% (12/23) in the control group (p =.01, 95% confidence interval of -58% to -4%), In patients >75 yrs of age, the mortality rate was 57% (12/21) in the protocol group and 61% (11/18) in the control group, Oxygen extraction ratios (O2ER) and; oxygen consumption values were sig significantly (p=.02) lower in the age >75 yrs group compared with the age 50 to 75 yrs group, Conclusions: Patients 50 to 75 yrs of age receiving a Do, of >600 mL/min/m(2) demonstrated a statistically significant (p =.01) improved survival rate over patients in the control group, Patients >75 yrs of age demonstrated no benefit from attempts to increase bo, to >600 mL/min/m(2), and they may have been overtreated as reflected by the lower O2ER values in this age group, Treating to an O2ER that reflects a balance between oxygen consumption and (D) over dot o(2) may be an alternative goal that allows individual titration.