EXPERIMENTAL models of gallstone formation have been described in many species1,2, but the lithogenic methods used were unphysiological or unnatural. In dogs, spontaneous gallstones are rare1, and human cholesterol gallstones placed in the dog gallbladder will dissolve3. Canine gallstones occur in biliary fistulas4, but the only report of gallbladder stones followed combined cholesterol-feeding and occlusion of the cystic duct5. To induce stones physiologically in this resistant species, we designed a diet in 1965 which hypothetically would alter bile salt–phospholipid–cholesterol–pigment relationships in bile and cause bile stasis as follows: low amounts of protein, high content of sugar and no added bile salts to reduce total bile salt and volume-flow of bile6,7; low content of total and unsaturated fat to reduce and alter phospho-lipids and bile salts in bile and to decrease gallbladder contractions8–10; and added cholesterol to overcome the cholesterol-holding capacity of dog bile1,11. Cornstarch and multiple small feedings were added further to simulate diets of certain human populations with a high incidence of cholelithiasis, for example, the Pima Indians12, or produce biliary stasis1,8. The diet was compounded of purified casein 10 per cent, sucrose 50 per cent, cornstarch 26 per cent, animal lard 5 per cent, cholesterol USP 1 per cent, salt mixture USP XIV 5 per cent, ‘Vitamin Diet Fortification Mixture’ 2 per cent and cellulose as filler 1 per cent (Nutritional Biochemicals Corporation, Cleveland, Ohio).