Oxidative stress and oxidative damage to tissues are common end points of chronic diseases, such as atherosclerosis, diabetes, and rheumatoid arthritis. The question addressed in this review is whether increased oxidative stress has a primary role in the pathogenesis of diabetic complications or whether it is a secondary indicator of end-stage tissue damage in diabetes. The increase in glycoxidation and lipoxidation products in plasma and tissue proteins suggests that oxidative stress is increased in diabetes. However, some of these products, such as 3-deoxyglucosone adducts to lysine and arginine residues, are formed independent of oxidation chemistry. Elevated levels of oxidizable substrates may also explain the increase in glycoxidation and lipoxidation products in tissue proteins, without the necessity of invoking an increase in oxidative stress. Further, age-adjusted levels of oxidized amino acids, a more direct indicator of oxidative stress, are not increased in skin collagen in diabetes. We propose that the increased chemical modification of proteins by carbohydrates and lipids in diabetes is the result of overload on metabolic pathways involved in detoxification of reactive carbonyl species, leading to a general increase in steady-state levels of reactive carbonyl compounds formed by both oxidative and nonoxidative reactions. The increase in glycoxidation and lipoxidation of tissue proteins in diabetes may therefore be viewed as the result of increased carbonyl stress. The distinction between oxidative and carbonyl stress is discussed along with the therapeutic implications of this difference.