Validated methods of determining murine body composition are required for studies of obesity in mice. Dual-energy X-ray absorptiometry (DEXA) provides a noninvasive approach to assess body fat and lean tissue contents. Similar to DEXA analyses in other species, body fat measurements in mice show acceptable precision but suffer from poor accuracy. Because fat and lean tissues each contain various components, these inaccuracies likely result from selection of inappropriate calibration standards. Analysis of solvents showed that the PIXImus2 DEXA gave results consistent with theoretical calculations. Male mice weighing 26-60 g and having body fat percentages ranging from 3 to 49% were analyzed by both PIXImus2 DEXA and chemical carcass analysis. DEXA overestimated mouse fat content by an average of 3.3 g, and algorithms were generated to calculate body fat from both measured body fat values and the measured ratio of high- to low-energy X-ray attenuations. With calibration to mouse body fat content measured by carcass analysis, the PIXImus2 DEXA gives accurate body composition values in mice.