Several different modifications of LDL have been described that convert it to a form recognized by one or more macrophage scavenger receptors. Modifications that can favor foam cell formation in vitro include oxidation, aggregation, enzymatic modification, complexing with immunoglobulins, and possibly others. 27 The best studied of these and the only one for which there is good in vivo data is oxidative modification. As reviewed elsewhere, 12 six different antioxidant compounds (probucol, probucol analogues, vitamin E, coenzyme Q, diphenylphenylenediamine, and butylated hydroxytoluene) have been studied in four different animal models of atherosclerosis (rabbits, mice, hamsters, and monkeys) and most of the results have been strikingly positive. A number of important ancillary lines of evidence are consonant with the hypothesis, including the fact that oxidation of LDL has been shown to occur in vivo and that OxLDL is demonstrable in lesions; that autoantibodies are generated against OxLDL and that the titers are correlated with the extent of atherosclerosis; that knocking out scavenger receptors (either scavenger receptor A or CD36) ameliorates atherosclerosis (establishing that some modified form of LDL is involved); and that knocking out 12/15-lipoxygenase, shown previously to be able to oxidize LDL, ameliorates atherosclerosis. 12, 15, 18–21, 28, 29 Because the basic pathology and pathogenesis of human atherosclerosis appears to be remarkably similar to that in these animal models, it would be surprising if oxidation were relevant to one and not the other. Why then have the clinical trials to date been mostly negative? One possibility is that oxidation is relevant to atherosclerosis in animal models but not in humans. However, the other possibility is that the design of current clinical trials has not been appropriate for testing the hypothesis.