Livestock producers would benefit from more precise predictions of the growth response from nutrients consumed. Previously published models are often limited by the realities of data collection and are unable to account for alterations to body composition, due in part to the response of visceral organs to an alternate diet. The computerized tomography (CT) scanning of lambs enables the analysis of changes in body composition of individual animals over time, potentially supporting better model development and testing. The aim of this experiment was to develop a repeatable method for the analysis of live lamb body composition using CT scans. A secondary aim was to compare the data collected from CT scanning during a feeding trial to 2 predictive lamb growth models. Cross-bred lambs were fed 2 feeding levels at 2 stages of maturity, with CT scans at the beginning and end of each 8-wk feeding period. The CT scan-derived values for body composition taken at the beginning of feeding periods were used as inputs for 2 existing lamb growth models. Predictions of body composition were compared with CT scan-derived values at the end of feeding periods. The CT scan analysis method used a proportion of images from each lamb to reduce manual image editing. The method was developed by comparing the estimated mass and volume of empty body components using all available CT scans to estimated values using a reduced number of scans from 12 lambs. The CT scan-derived lean tissue mass aligned with model predictions at the end of each feeding period, however, CT scan-derived fat mass was greater than predictions by both models especially for the high feeding level at the later stage of maturity. These results highlight that the analysis of body composition using CT scans requires further validation, particularly for the viscera, and that models likely require refinement to better predict the efficiency of energy utilization by different tissues. The use of live animal CT scans can provide more accurate predictions of the growth of saleable products than measuring liveweight alone and will enable ruminant growth models to better adapt to different genetics and changing diets than comparative slaughter. To replicate the current data using comparative slaughter would require 4 times the animals, as individual lambs were CT scanned 4 times in this study, demonstrating the potential value of CT scanning in live animal research.