The aim of this study was to estimate correlated responses in lean, fat and bone weights in vivo in Dorset Down sheep selected for lean tissue growth. Over the period 1986-1992 inclusive, the lean tissue growth line had been selected using two economic indices for an increased aggregate breeding value incorporating predicted lean and fat weights with positive and negative economic weightings, respectively. The control line was selected for no change in lean tissue growth each year. Animals were born and run on pasture all year round. X-ray computer tomography was used to estimate the weights of lean, fat and bone in vivo in the 1994-born sheep, aged 265-274 days and selected randomly into 12 rams and 12 ewes from the selected line and 10 rams and 9 ewes from the control line. The lean tissue growth line had significantly greater responses in lean weight (+0.65 Â± 0.10 kg) and lean percentage (+1.19 ± 0.17%) and significantly lesser fat weight (-0.36 ± 0.08 kg) and fat percentage (-1.88 ± 0.20%) compared to the control line. There was a significant increase in bone weight (+0.27 ± 0.03 kg) and bone percentage (+0.69 ± 0.09%) in the lean tissue growth line compared to the control line. Responses differed significantly between sexes of the lean tissue growth line, rams having a greater response in weight of lean (+1.22 ± 0.20 vs. +0.08± 0.22 kg) and bone (+0.45 ± 0.06 vs. +0.09 ± 0.07 kg), and a lesser response in weight of fat (-0.03 ± 0.15 vs. -0.70 ± 0.16 kg) than the ewes. Selection led to significant changes in lean (increase) and fat weights (decrease), and bone weight increased. Although responses in the lean tissue growth line differed significantly between sexes, there were confounding factors due to differences in management and lack of comparison at equal stage of development. Therefore, to assess real genetic differences further studies should be conducted taking these factors into consideration.