AbstractMetal catalysis of membrane lipid oxidation has been thought to occur only at cell surfaces. However, conflicting observations of the pro‐oxidant activity of ferric (Fe3+) vs ferrous (Fe2+) forms of various chelates have raised questions regarding this dogma. This paper suggests that the solubilities of iron complexes in lipid phases and the corresponding abilities to initiate lipid oxidation there, either directly or via Fenton‐like production of reactive hydroxyl radicals, are critical determinants of initial catalytic effectiveness.Partitioning of Fe3+ and Fe2+ complexes and chelates into bulk phases of purified lipids was quantified by atomic absorption spectroscopy. mM solutions of iron salts partitioned into oleic acid at levels of about micromolar. Ethylenediamine tetraacetic acid (EDTA) and diethylenetriaminepentaacetic acid (DTPA) chelates were somewhat less soluble, while adenosine diphosphate (ADP) chelates, and ferrioxamine were soluble as chelates at greater than 10−5 M. Solubilities of all iron compounds in methyl linoleate were 10‐ to 100‐fold lower.To determine whether Fenton‐like reactions occur in lipid phases, H2O2 and either Fe2+ or Fe3+ and a reducing agent were partitioned into the lipid along with the spintrap 5,5‐dimethyl‐1‐pyrroline‐N‐oxide (DMPO), and free radical adducts were recorded by electron paramagnetic resonance (EPR). Hydroxyl radicals (OH.) adducts were observed in oleic acid, but in lipid esters secondary peroxyl radicals predominated, and the presence of OH. adducts was uncertain.