Methyl tert-butyl ether which is a powerful cholesterol monohydrate solvent does not completely dissolve mixed cholesterol gallstones when directly infused into the biliary tree. In this work, we compared the effect of various solvents containing different proportions of methyl tert-butyl ether and dimethylsulfoxide in anhydrous and aqueous systems on the in vitro solubilization of human cholesterol stones. The dissolution rates of cholesterol obtained in the presence of methyl tert-butyl ether was markedly decreased when 10 p. 100 water was added. In contrast, the addition of dimethylsulfoxide (30 p. 100) to methyl tert-butyl ether-water system enhanced the stone-solvent contact, improved the cholesterol dissolution rates and left less stone debris. A subsequent dissolution with an alkaline, pH = 8.8, aqueous dimethylsulfoxide-ethylenediaminetetraacetic acid solution strongly reduced the non cholesterol residues. In vivo, nearly complete dissolution of human cholesterol stones implanted in the gallbladder of rabbits was obtained within 8 hours when methyl tert-butyl ether/dimethylsulfoxide (70/30) solvent was infused at a rate of 0.6 ml/h/kg. With methyl tert-butyl ether, only 84 p. 100 of the original stone weight was dissolved. The infusion of these solvents leads to morphological changes in the gallbladder wall with some focal ulcerations. These alterations can be almost completely recovered after two weeks. No histologic evidence of hepatic, duodenal or renal damage was found. We conclude that the mixture methyl tert-butyl ether/dimethylsulfoxide (70/30) constitutes a good solvent for mixed cholesterol stones. Compared with pure methyl tert-butyl ether, the mixed system allows for a more rapid and a more complete dissolution of gallstones.