Recent studies have suggested that free radical scavenger administration reduces the rate of development of diaphragm fatigue. Much of this work has been done, however, using in vitro muscle preparations; the purpose of the present study was to assess the effect of scavengers on in vivo diaphragm contractile function. To accomplish this, we compared the rate of development of fatigue of the electrically stimulated diaphragm in four groups of dogs: (1) animals given intravenous polyethylene glycol adsorbed superoxide dismutase (PEG-SOD, 2,000 units/kg) 1 h before a fatigue trial; (2) a group given intravenous dimethylsulfoxide (DMSO, 0.5 ml/kg of a 50% solution) before fatigue; (3) a group given saline before fatigue; and (4) a group treated with denatured PEG-SOD (2,000 units/kg) before fatigue. We measured diaphragmatic concentrations of thiobarbituric acid reactive substances (TBAR), a marker of free radical-mediated lipid peroxidation, on muscle samples taken at the conclusion of fatigue trials. As a control, we also measured TBAR concentrations for muscle samples taken from nonfatigued diaphragm. We found that the rate of development of diaphragm fatigue was much greater in saline and denatured PEG-SOD-treated groups than for animals pretreated with either PEG-SOD or DMSO, with force falling to 23 +/- 4, 21 +/- 4, 50 +/- 7, and 47 +/- 6% of its initial value, respectively, over a 2-h period of electrophrenic stimulation in these four groups of animals (p < 0.01). TBAR concentrations in fatigued diaphragm from saline and denatured PEG-SOD-treated animals were significantly higher than levels for either nonfatigued fresh diaphragm or fatigued diaphragm taken from PEG-SOD- or DMSO-treated animals (p < 0.01). These data suggest that diaphragm fatigue resulting from repetitive low-frequency stimulation is associated with lipid peroxidation within this muscle and that pretreatment with free radical scavengers prevents lipid peroxidation and reduces the rate of development of fatigue.