The ability of mice deficient in galactose-1-phosphate uridyltransferase (GALT) to metabolize galactose was determined in animals weaned to a mouse chow diet for a 4-wk period. When given [14C]galactose intraperitoneally, these animals slowly oxidized the sugar, excreting only 5.5% of the dose as 14CO2 in 4 h, whereas normal animals excreted 39.9%. These results mimic those seen in human galactosemic patients given isotopic galactose. When given 10 micromol of [1-13C]galactose, normal animals excrete small amounts of labeled galactose and galactonate but no galactitol in urine whereas GALT-deficient mice excrete significant amounts of all of these as labeled compounds in urine. When challenged with galactose, only about 20% of the dose is excreted in urine, and even on the chow diet, significant amounts of galactose, galactonate, and galactitol are excreted in urine. These compounds are also found to be present in liver, kidney, and brain, except that galactonate is not found in brain. Galactose-1-phosphate accumulates in red blood cells to levels found in humans exposed to large amounts of galactose, and galactose-1-phosphate is found in increased amounts in liver, kidney, and brain of GALT-deficient animals. There was no difference in the hepatic concentration of uridine diphosphate galactose and uridine diphosphate glucose between normal and GALT-deficient mice. The explanation for the presence of galactose and its conversion products in tissues and urine of affected mice appears to be related to the presence of approximately 1.75% of galactose-containing carbohydrates in the chow, which becomes bioavailable to mice. Despite the presence of galactose and its metabolites in tissues and urine and impaired ability to oxidize the sugar, the GALT-deficient animals are indistinguishable from normal animals and do not exhibit the phenotype of humans with GALT-deficiency galactosemia.