Apparently favorable effects of antiatherosclerosis drugs as assessed by changes in surrogate markers of cardiovascular disease risk are frequently relied upon for drug approval and labeling. Surrogates must be biologically plausible and adequately validated but are, by definition, imperfect as predictors of ultimate outcome (i.e., serious morbidity and mortality). Surrogate markers utilized in the study of drugs for the treatment of atherosclerotic cardiovascular disease may be classified as laboratory/biochemical, anatomic/morphologic, and functional. The places for various surrogates in all three categories in the development of lipid altering drugs are discussed.