1. Studies of tissues distribution in beef carcasses of East African ‘Boran’ cattle revealed very high correlations between the separable tissue analysis of the full tenth rib sample joint and of the full side.

2. A comparative study of tenth rib sample joints, from twelve steers and five cows cut at various positions along the length of the rib showed that the full rib sample gave a higher correlation (r = 0·948), with the full side in respect of all tissues except the subcutaneous fat, than any other sample joint. The subcutaneous fat of the whole side could not be predicted with reliability by the use of any rib sample, the highest correlation (r = 0·558) existing between the half-rib sample and the full side.

3. From detailed carcass studies of thirty-two steers and eight cows, including the twelve steers and five cows mentioned above, prediction equations are presented for estimating the percentages of lean, total fat, and butchers' bone in whole carcasses from single complete tenth rib samples. Standard deviations being 1·53% for lean, 1·34% for total fat in steers, 1·13% for total fat in cows and 0·88% for butchers' bone.

4. Where X1, X2 and X3 are respectively the percentages of lean, total fat and bone in the sample joint the prediction equations recommended are as follows:

For the prediction of lean

(i) In carcasses of either sex.

(f) Percentage of lean in the carcass = 0·431X1 + 0·475X3 + 33·58 (subject to a standard deviation of 1·53%).

For the prediction of total fat

(i) In steer carcasses.

(d) Percentage total fat in carcass = 0·664X2 − 0·78 (subject to a standard deviation of 1·34%).

(ii) In cow carcasses.

(e) Percentage of total fat in carcass = 0·641X2 + 0·23 (subject to a standard deviation of 1·13%).

For the prediction of bone

(i) 100 −(equation (f) +equation (d) or (e)).

5. A practical test of the equations referred to in the text is given in Table 4.