Four 99mTc and three 123I labeling methods were evaluated for their suitability to label low density lipoproteins (LDL) for the purpose of scintigraphic biodistribution studies. For 99mTc these methods were: direct incorporation in LDL of 99mTcO4- using sodium dithionite (dithionite method); a method using first N,N-dimethylformamide to prepare a 99mTc-complex reacting with LDL in a subsequent step (DMF method); a technique in which 99mTcO4- is first coupled to a diamide dithiolate derivative of pentanoic acid by reduction with dithionite, followed by coupling of this ligand to LDL (N2S2 method); and a method using sodium borohydride and stannous chloride as reducing agents (borohydride method). The iodination techniques were based on oxidation of I-→I+, using iodine monochloride (ICl method), 1,3,4,6-tetrachloro-3,6-diphenylglycoluril (Iodogen method), and N-bromosuccinimide (NBS method) as oxidants. We studied labeling yields, modification of LDL caused by the labeling procedures using agarose-gel electrophoresis, and radiochemical stability of the labeled LDL complex upon incubation in plasma at 37°C for 15 h. We used Sepharose CL6B chromatography to separate LDL from other plasma proteins. We also examined whether LDL isolated from frozen plasma (Pool-LDL) gave results similar to LDL obtained from freshly prepared plasma (Fresh-LDL). Pool-LDL radiolabeled by the dithionite, DMF, NBS, and Iodogen methods lost its label upon incubation with plasma. This also happened with Fresh-LDL when the DMF, NBS and Iodogen methods were used. Upon agarose-gel electrophoresis, no modification of LDL was observed with all methods when the radionuclide/LDL ratio was kept low. However, when higher ratios were used, the LDL was detectably modified by the DMF and Iodogen methods as evidenced by its increased electronegativity. For both Pool-LDL and Fresh-LDL, 99mTc-labeled LDL prepared by the N2S2 and borohydride method, and 123I-labeled LDL as obtained by the ICl method were both stable and apparently unmodified. For the dithionite method this was also true when Fresh-LDL was used. The plasma clearance studies of these stable radiolabeled LDL preparations in rabbits showed similar clearance rates for 99mTc-labeled LDL as obtained by the N2S2 method and 125I-labeled LDL as iodinated by the ICl method, for both Fresh- and Pool-LDL. 99mTc-labeled Fresh-LDL produced by the dithionite method showed a slightly accelerated plasma decay, whereas both Fresh- and Pool-LDL labeled by the borohydride method were cleared significantly faster from the circulation. The better labeling results of Fresh-LDL as compared with Pool-LDL, and possible modification of Pool-LDL induced during storage, suggests that the use of Fresh-LDL is preferable. The above data indicate that only Fresh-LDL labeled by the dithionite method, the N2S2 method, and the ICl method are potentially acceptable as scintigraphic agents in biodistribution studies.