Exogenous bovine pituitary somatotropin (GH) can influence markedly body composition in fattening lambs. However, neither the effects of biosynthetic somatotropin nor the effects of dose and method of administration have been reported. Fifty Dorset-cross lambs (female and castrated male) were given concentrate ad libitum and treated between 10 and 22 weeks of age with biosynthetic bovine somatotropin either dissolved in buffer and injected subcutaneously (s.c.) in proportion to body weight (0·025, 0·1 or 0·25 mg/kg per day), dissolved in buffer and continuously infused s.c. (0·1 mg/kg per day) or suspended in olive oil and injected s.c. (0·1 mg/kg per day), and compared with 10 untreated control lambs. Somatotropin had little effect on live-weight gain (controls = 228 g/day; final live weight 37 kg), food intake and food conversion efficiency, and only marginally increased the weight of muscle and bone dissected from the shoulder joint. The weights of the major fat depots in the abdominal cavity and of fat dissected from the shoulder joint were linearly related to dose of somatotropin (P < 0·001). Lambs given the highest dose had less visceral fat (1·18 v. 2·84 kg; P < 0·001) and proportionately less fat (285 v. 374 g/kg; P < 0·001) and more muscle (542 v. 447 g/kg) and bone (172 v. 149 g/kg; P < 0·001) in the shoulder joint than control lambs. There was a positive curvilinear relationship (P < 0·01) between clean wool growth (mid-side patch sample) and dose; lambs injected daily with 0·1 mg somatotropin per kg grew one-third more wool than control lambs. Method of administration affected plasma somatotropin profiles but had no significant influence on any of the responses measured. The anabolic actions of somatotropin may have been limited in this experiment by a low sodium concentration in the diet. It is concluded that the lipolytic/anti-lipogenic effect of somatotropin can occur in the absence of conditions conducive to the stimulation of muscle protein deposition and that the response is dose dependent but not influenced by method of administration of the hormone.