Lipid peroxidation is thought to be a major contributing factor in neurological injury following cardiac arrest. Because iron availability is a prerequisite for lipid peroxidation, this experiment was designed to examine the natural time course of iron release, lipid peroxidation, and cerebral polyunsaturated fatty acid content following a 15-minute cardiac arrest in dogs. Large mongrel dogs were anesthetized with ketamine and halothane and divided into three groups of five each. In two groups, cardiac arrest was induced with KCl. After 15 minutes of cardiac arrest, the dogs were resuscitated by five minutes of internal cardiac massage, epinephrine, bicarbonate, and internal defibrillation. All ten dogs were resuscitated and supported by a standard intensive care protocol until tissue harvest. A 3-g portion of parietal cerebral cortex was obtained from the nonischemic dogs (n = 5), or at two hours (n = 5), or eight hours (n = 5) after resuscitation. Total tissue iron was measured by an atomic emission spectrometer; low molecular weight species (LMWS) iron by the o-phenanthroline test on an ultrafiltered sample; and lipid peroxidation by both the thiobarbituric acid test (TBARS) and determination of the tissue content of lipid double bonds, calculated by first fractionating the lipids by gas-liquid chromatography and then measuring the double bonds in each fraction by spectrometry and summing the results. Univariate ANOVA demonstrated all variables except total tissue iron to have significance at P less than .02. At two hours of reperfusion, LMWS iron and TBARS were significantly elevated above nonischemic control levels.(ABSTRACT TRUNCATED AT 250 WORDS)