We evaluated the safety and efficacy of Fluosol DA-20% (FDA) as a blood substitute in the treatment of severe anemia. Thirty-six patients received either FDA (n = 21) or crystalloid/hydroxyethyl starch (CHS) (n = 15) as part of a randomized, controlled trial. Ten patients received FDA as part of a humanitarian protocol. All were Jehovah's Witnesses who refused transfusion, had bled recently, and had average Hgb levels of 4.3 g/dl. After pulmonary artery catheter insertion, each patient was infused with CHS to attain a pulmonary artery wedge pressure (WP) of 10 to 18 mm Hg. FDA was given as a one-time dose of 30 ml/kg. Data were collected at baseline, 12, 24, and 48 h. None of the patients with negative reactions to a 0.5-ml test dose of FDA had adverse reactions to the subsequent infusion. The plasma or dissolved component of oxygen content was significantly higher in the FDA group at 12 h (FDA group 1.58 +/- 0.47 ml/dl, control group 1.01 +/- 0.31 ml/dl, p less than .02, t-test). Nineteen patients died: 12 (37.5%) FDA, seven (46.6%) control. The difference was not significant. We conclude the following: a) FDA can be given safely to severely anemic patients in doses of 30 ml/kg; b) FDA significantly increased the dissolved component of oxygen content after 12 h but the effect did not persist; c) severely anemic patients can survive without transfusion although mortality is high. In this study, inability of FDA to sustain increased oxygen content was due in part to the rapid elimination of FDA and also to the limited amount given.(ABSTRACT TRUNCATED AT 250 WORDS)