OBJECTIVE The aim of this study was to compare the metabolic effects of GH administered subcutaneously either once or twice daily. The actions of GH might depend upon a pulsatile pattern of serum GH. Pulsatile and continuous intravenous delivery of GH, however, induce similar short-term metabolic effects in GH deficient patients. An improved growth response is obtained in GH deficient children when a fixed weekly GH dose is administered by daily subcutaneous injections instead of twice or thrice-weekly intramuscular injections. A more pulsatile pattern and serum GH levels above zero might be achieved by further increasing the injection frequency. Increased daytime GH levels might, however, adversely affect the circadian patterns of metabolic indices, which have been demonstrated to be more successfully reproduced by evening compared with morning GH administration. DESIGN AND MEASUREMENTS In a cross-over study, 8 GH deficient patients (age 16-43 years) were treated with 3 IU/m2/24 h of human GH. The dose was injected in the evening for 4 weeks and for another 4 weeks two-thirds was injected in the evening and one-third in the morning. At the end of each period the patients were admitted to the hospital for 37 hours. Steady-state profiles of GH, IGF-I, IGF binding proteins 1 and 3, insulin, glucose, lipid intermediates and metabolites were obtained following administration of 3 IU/m2 of GH (at 1900 h (one injection) and at 1900 and 0800 h (two injections)). RESULTS Similar mean integrated levels of serum GH (mU/l) were obtained (7.46 +/- 0.84 (one injection) vs 6.46 +/- 0.62 (two injections) (P = 0.15)). Mean levels +/- SEM of serum IGF-I (micrograms/l) were significantly increased (P < 0.01) following two daily GH injections (330.3 +/- 48.1 (one injection) vs 399.1 +/- 53.0 (two injections)). Serum IGFBP-3 levels were not significantly different on the two occasions, while levels of the GH independent IGFBP-1 (micrograms/l) were slightly but significantly lower following twice-daily GH injections (1.61 +/- 0.42 vs 1.13 +/- 0.56, respectively (P < 0.04)). The pattern of IGFBP-1 was opposite to that of insulin. Similar levels of insulin and glucose were obtained with both GH regimens, while levels of non-esterified fatty acids were significantly higher following once-daily GH injection (P < 0.001). CONCLUSIONS Twice-daily GH injections, apart from producing a more physiological serum GH profile, were superior to one injection in increasing serum IGF-I and decreasing IGFBP-1 levels. Both of these changes tend to amplify the effects of the administered GH. Twice-daily injections, however, resulted in lower night-time levels of lipid intermediates.