SUMMARY Twenty-three Poll Hertford half-sib steers were available for experiment. Eighteen of these were 19 months old, and 5 were 31 months old. At the commencement of the experiment all the older steers, and 5 of the younger steers were slaughtered and dissected as control groups (C1 and C2). The remaining 13 young steers were allotted to experimental groups. Data from the 2 control groups and the differences between them thus provided basic lines of uninterrupted growth for comparison with the experimental groups. A semi-starved group (S) was slaughtered after 66 days of treatment, and a recovered group (R) after 207 days of high-plane nutrition following treatment similar to that of group S. The percentage increases of the liveweight, carcase weight, non-carcase body weight, muscle weight and bone weight were all of similar order of approximately 25% between groups C1 and C2, whereas dissectible fat increased by 70.4%. During semi-starvation dissectible fat, and particularly subcutaneous fat, suffered the greatest percentage decrease (69.7%), exceeding carcase weight (22.0%), dissectible muscle (20.9%), liveweight (15.7%), non-carcase body weight (8.8%) and bone (1.7%). The actual weight of muscle lost was 1.18 times greater than the weight of fat lost. During recovery the percentage increase of non-carcase body weight was greatest (46.7%) followed by dissectible muscle (42.6%), liveweight (40.9%), carcase weight (35.5%), dissectible fat (30.9%) and bone (9.0%). The composition of the weight gain during recovery included 2.8 parts of muscle to each one part of dissectible fat, compared with 1.71 parts of muscle to fat during the uninterrupted growth between the control groups. The weight of the late developing muscles was depleted most, and that of the early developing muscles least, during semi-starvation. The muscle-weight distribution of the recovered steers did not differ from the controls, which illustrated the ability of the musculature to recover. The commercial dressed carcases of the recovered cattle contained a higher proportion of muscle than that expected from normally fed cattle of that weight and age; however, due to lower dressing percentage the actual muscle weight was lower than the expected value for cattle fed on an uninterrupted plane of nutrition. It is suggested that cattle which undergo semi-starvation at more mature ages have a larger proportion of muscle to fat in their weight gain during recovery than have calves submitted to similar treatment. The present study indicates that following periods of semi-starvation, early-maturing types of cattle may produce carcases with more optimum fat cover at the light carcase weights suitable for certain markets. The need for critical examination of the effects of semi-starvation and recovery on different types of cattle is stressed.