Commonly used spectrophotometric methods for determining the extent of lipid peroxidation in animal tissue extracts, such as measurements of diene conjugation and thiobarbituric acid reactive substances (TBARS), have been criticized for their lack of specificity. This study shows that lipid hydroperoxides can be effectively quantified in animal tissue extracts using an assay based on the formation of a Fe(III)xylenol orange complex. Addition of H2O2, cumene hydroperoxides, or methanolic tissue extracts to an acidic reaction mixture containing 0.25 mM Fe(II) and 0.1 mM xylenol orange caused the formation of a broad Fe(III)xylenol orange complex absorbance peak at 560-580 nm with a corresponding decrease in the xylenol orange peak at 440 nm. Complex formation measured at 580 nm was saturable with both xylenol orange and Fe (II) concentration. Addition of ascorbic acid, GSH, and cysteine (0.3-5 mM) caused a saturable reduction of the Fe(III)xylenol orange complex. Formation of the Fe(III)xylenol orange complex was linear with the amount of tissue extract added. A significant correlation (r = 0.88, p < 0.005) existed between the xylenol orange method of estimating lipid peroxidation and the conventional TBARS assay in a series of animal tissues tested. The time course of increase in A580nm in tests using tissue extracts was typical of a free radical reaction; a lag phase was followed by a log phase. No increase in A580nm was observed up to 24 h when highly peroxidizable arachidonic acid was assayed. These results indicate that the formation of the Fe(III)xylenol orange complex reflects a chemical amplification of the original level of lipid hydroperoxides present in tissue extracts and that peroxidizable lipids do not influence the assay. The potential usefulness of the xylenol orange assay for comparative biochemical and toxicological studies of oxidative stress is discussed.