Ten adult patients (three male, age range, 15 to 67 years) with established ARDS were studied for serial changes in the proteinaceous antioxidant activities of their plasma. All had LISs in excess of 2.5 on admission to the study. Blood samples were taken as soon as possible after the diagnosis of ARDS, and from 10 patients at risk of developing ARDS. These were compared with healthy control subjects. Deoxyribose, phospholipids, and DNA were used as markers of damage in reactions generating inorganic and organic oxygen radicals and an oxo-iron species. The ability of plasma to inhibit such damage was expressed as antioxidant activity. This study does not address the clinical problem of why certain "at risk" patients develop ARDS, but rather the question of why some patients with established ARDS are better able than others to survive the disease. ARDS patients had transferrin levels that were significantly lower (1.76 +/- 0.13 gm/L) than those of normal controls (2.91 +/- 0.12 gm/L, p < 0.001), which decreased the ability of their plasma to protect phospholipid membranes and DNA against iron-stimulated free radical damage. The iron-oxidizing antioxidant properties of plasma were mainly dependent on the protein ceruloplasmin, concentrations of which were significantly higher in ARDS patients (0.387 +/- 0.04 gm/L) than in healthy controls (0.265 +/- 0.03 gm/L, p = < 0.05) or patients at risk of ARDS (0.24 +/- 0.04 gm/L, p = < 0.05). The iron-oxidizing (ferroxidase) antioxidant activities of plasma from ARDS patients, however, were similar to those of both control groups. Measurement of plasma ferroxidase activities confirmed that although more ceruloplasmin was present in the plasma of ARDS patients, enzyme activities were comparable to those of both control groups, which was suggestive of a loss of ceruloplasmin ferroxidase activity. Scavenging and radical-stimulating properties of plasma (devoid of iron-binding and iron-oxidizing properties) was partly dependent on protein thiol groups, which were lower in ARDS patients and in patients at risk of developing ARDS. Serial sample analysis revealed that ARDS patients showed substantial daily variations in biochemical parameters, implying that single time-point sampling may be unsuitable when studying these patients.