Reductive methylation of at least 30% of the lysine residues or 1,2-cyclohexanedione modification of 45% of the arginine residues prevented low density lipoproteins (LDL) from binding to cell surface receptors of fibroblasts in vitro, without significantly altering other physical or chemical properties of the LDL. When rat or human LDL with more than 30% of the lysine residues methylated were injected intravenously into rats, the clearance of these lipoproteins from the plasma was slowed considerably. The half-life of the reductively methylated LDL was approximately twice that obtained for control (unmodified) LDL, and the value for the fractional catabolic rate was approximately half that of the control. Furthermore, when human LDL modified by reductive methylation were injected into rhesus monkeys, the rate of clearance was similarly retarded, and the value for the fractional catabolic rate was reduced by approximately 50% as compared with the value for control LDL. A dual isotope labeling technique ((125)I and (131)I) was used to compare the disappearance of the control and modified LDL in the same animal. It was demonstrated that not only modification of lysine residues but also modification of the arginine residues with 1,2-cyclohexanedione retarded the plasma clearance of the rat LDL. However, the cyclohexanedione modification was spontaneously reversible at 37 degrees C, whereas reductive methylation of the lysine residues was stable. It is concluded that the selective chemical modification of lysine or arginine residues of LDL interferes with the normal uptake of these lipoproteins in vivo as well as by fibroblasts in vitro. These data provide an estimation of the level of receptor-mediated clearance of LDL from the plasma, a value that may be as high as 50% in rats and monkeys.