The effects of two methods of preparing small unilamellar vesicles (SUV) (detergent removal or sonication) on their in vivo elimination and tissue distribution was investigated in rats. The SUV prepared by either method had the same size distribution and lipid composition (egg yolk phosphatidylcholine/cholesterol/dipalmitoyl phosphatidylethanolamine or palmitic acid = 20/10/0.3, molar ratio). Three types of SUV made by either method were prepared. These contained one of three different surface ligand-binding functional groups (N-hydroxysuccinimide ester of palmitic acid, NHSP; glutaraldehyde-phosphatidylethanolamine, GA-PE; N-[4-(p-maleimidophenyl)butyryl]phosphatidylethanolamine, MPB-PE). SUV prepared by detergent removal were eliminated slowly from the circulation, and exhibited a low liver uptake and little leakage of [3H]inulin. There was no significant difference in elimination of the NHSP-SUV, GA-SUV or MPB-SUV prepared by detergent removal and their tissue distribution was similar. In contrast, the sonicated SUV were eliminated from the circulation much more rapidly mainly by liver uptake. The leakage of [3H]inulin from sonicated SUV into urine was relatively large. When sonicated control-SUV were prepared in the presence of the antioxidant, alpha-tocopherol (alpha-T-SUV), which reduces lipid peroxidation during sonication, the alpha-T-SUV were eliminated slowly with only a low liver uptake. Our results indicate that the rapid elimination and greater liver uptake of sonicated SUV is partly due to lipid peroxidation during preparation. These findings have relevance to the use of liposomes as a drug delivery system.