High-resolution microcomputed tomography technology has allowed researchers to use live mice to address questions that previously could be answered only at necropsy. Serial analyses of the same mouse allow tissue changes to be followed over time. The ability to follow a single mouse noninvasively can decrease the total number of mice required for the study. The magnitude of inter-mouse variation for matched mice undergoing microcomputed tomography has not been determined previously. We selected lung and contrast-enhanced stomach as tissues of standard size and anatomical structure that were hypothesized to vary minimally between mice. The analyses of the tissue volumes from matched mice showed considerable variation among mice, among multiple sequential scans of the same mouse, and even among multiple evaluations of the same scan. More variation occurred with repeated scans of the same mouse (intramouse variation) than between mice (intermouse variation). In addition, significant variation and obvious bias was detected between the 2 scan evaluators. These data suggest that to obtain the widest range of possible values, among which the true value would be found, multiple analyses of multiple scans of the same mouse must be performed by multiple scan evaluators.