Data from 54 hybrid (mainly Large White x Landrace) pigs (18 boars, 18 gilts, and 18 barrows) were used to quantify and mathematically describe the differential growth and development of body components of live pigs. The pigs were 32.4 +/- 3.2 kg of BW and 70 +/- 1 d of age (mean +/- SD) at the beginning of the study, were individually penned and fed ad libitum, and were weighed weekly. Computed tomography (CT) imaging was used to determine the weights of lean, fat, bone, and skin tissue in the live pig at 30, 60, 90, 120, and 150 kg of BW. For each target BW, the sum of all the weights of the body components, as assessed by CT, was referred to as CT BW. Linear and nonlinear models were developed to evaluate the patterns of growth and development of each body component relative to CT BW. The correlation between the actual BW and CT BW was close to unity (r = 0.99), indicating that CT scanning could accurately predict the BW of pigs. Across sex and castrate status, percentage of fat (fat weight/CT BW) in the pig was least (11.2%) at the 30-kg target BW and continued to increase to 22.6% by the 150-kg target BW. Percentage of lean, however, was greatest (67.2%) at the 30-kg target BW and continued to decrease to 53.4% by the 150-kg target BW. The sex or castrate status x target BW interaction was significant (P < 0.05) for all the body components, indicating that the developmental patterns were different among sex or castrate status. Barrows were fatter relative to gilts, which in turn were fatter than boars. For lean, the observed pattern for sex or castrate status differences was opposite that for fat. To predict responses to management strategies on growth and development in pigs, accurate mathematical models are required, and the results of this study indicate that the nonlinear (e.g., augmented allometric and generalized nonlinear) functions provided better descriptions of the growth and development of most body components of the live pig than did the simpler (e.g., linear and allometric) models.