We have shown that primates adequately compensate for acute normovolemic anemia to hematocrits (HCT) of 10%. We have described a whole-body extraction ratio (O2 consumption/O2 delivery; ER) of 50% as a reliable physiologic indicator of transfusion need. There is concern that whole-body ER may not accurately reflect impaired myocardial metabolism. The onset of significant lactate production by the left ventricle is an indicator of anaerobic metabolism. Our purpose is to compare left ventricular lactate metabolism (arterial-coronary sinus lactate; delta [L]) to ER in acute normovolemic anemia. Fourteen adult baboons were anesthetized, paralyzed, and ventilated on room air. Left atrial, coronary sinus, aortic, and Swan-Ganz catheters were inserted. Experimental animals (N = 7) were hemodiluted, at constant left atrial pressure (LAP), with 5% human serum albumin (HSA) to a HCT below 4%. Control animals (N = 7) underwent similar volume exchanges, also at constant LAP, with homologous RBCs resuspended in HSA. Whole-body extraction ratio and left ventricular lactate production were measured at baseline and at hematocrits of 20, 10, 6, and 4% in the experimental group. Data were obtained at similar volume exchange points in the control group. Significant lactate production occurred only in the experimental animals (P less than 0.05) when extraction ratio exceeded 50%. Significant lactate production does not occur before the whole-body ER exceeds 50%. ER appears to be a valid indicator of myocardial metabolism in anemia, in this setting.